[Buildroot] [PATCH 1/1] wpa-supplicant: Add NL80211 support option

[Buildroot] [PATCH 1/1] wpa-supplicant: Add NL80211 support option

[Jean-Baptiste Theou]

Signed-off-by: Jean-Baptiste Theou <jtheou@adeneo-embedded.us>
---
 package/wpa_supplicant/Config.in         | 6 ++++++
 package/wpa_supplicant/wpa_supplicant.mk | 6 ++++++
 2 files changed, 12 insertions(+)

diff --git a/package/wpa_supplicant/Config.in b/package/wpa_supplicant/Config.in
index 81c61ac..2fb20ab 100644
--- a/package/wpa_supplicant/Config.in
+++ b/package/wpa_supplicant/Config.in
@@ -22,6 +22,12 @@ config BR2_PACKAGE_WPA_SUPPLICANT_EAP
 	help
 	  Enable support for EAP.
 
+config BR2_PACKAGE_WPA_SUPPLICANT_NL80211
+	bool "Enable NL80211"
+	depends on BR2_PACKAGE_LIBNL
+	help
+	  Enable support for NL80211.
+
 config BR2_PACKAGE_WPA_SUPPLICANT_HOTSPOT
 	bool "Enable HS20"
 	help
diff --git a/package/wpa_supplicant/wpa_supplicant.mk b/package/wpa_supplicant/wpa_supplicant.mk
index 0ca2ce5..2807151 100644
--- a/package/wpa_supplicant/wpa_supplicant.mk
+++ b/package/wpa_supplicant/wpa_supplicant.mk
@@ -49,6 +49,12 @@ else
 	WPA_SUPPLICANT_CONFIG_DISABLE += CONFIG_EAP
 endif
 
+ifeq ($(BR2_PACKAGE_WPA_SUPPLICANT_NL80211),y)
+	WPA_SUPPLICANT_CONFIG_ENABLE += CONFIG_DRIVER_NL80211
+else
+	WPA_SUPPLICANT_CONFIG_DISABLE += CONFIG_DRIVER_NL80211
+endif
+
 ifeq ($(BR2_PACKAGE_WPA_SUPPLICANT_HOTSPOT),y)
 	WPA_SUPPLICANT_CONFIG_ENABLE += CONFIG_HS20 \
 		CONFIG_INTERWORKING
-- 
2.0.3

[Buildroot] [PATCH 1/1] wpa-supplicant: Add NL80211 support option

[Gustavo Zacarias]

On 08/07/2014 05:28 PM, Jean-Baptiste Theou wrote:

> Signed-off-by: Jean-Baptiste Theou <jtheou@adeneo-embedded.us>
> ---
>  package/wpa_supplicant/Config.in         | 6 ++++++
>  package/wpa_supplicant/wpa_supplicant.mk | 6 ++++++
>  2 files changed, 12 insertions(+)

Hi.
Hmmm why?
This is done automatically if the libnl package is selected/around.
We really don't want to thread too thin into hostapd & wpa_supplicant
config options.
Regards.

[Buildroot] [PATCH 1/1] wpa-supplicant: Add NL80211 support option

[Jean-Baptiste Theou]

HI,

Ok, please discard the patch.

IMHO, wpa_supplicant should provide the support of NL80211 by default 
('depends on BR2_PACKAGE_LIBNL) since it's the new standard.

I wasn't expected to have to select LIBNL manually.

Best regards,
On 08/07/2014 01:37 PM, Gustavo Zacarias wrote:
> On 08/07/2014 05:28 PM, Jean-Baptiste Theou wrote:
>
>> Signed-off-by: Jean-Baptiste Theou <jtheou@adeneo-embedded.us>
>> ---
>>   package/wpa_supplicant/Config.in         | 6 ++++++
>>   package/wpa_supplicant/wpa_supplicant.mk | 6 ++++++
>>   2 files changed, 12 insertions(+)
> Hi.
> Hmmm why?
> This is done automatically if the libnl package is selected/around.
> We really don't want to thread too thin into hostapd & wpa_supplicant
> config options.
> Regards.
>
>

-- 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140807/d41578fa/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.jpg
Type: image/jpeg
Size: 8458 bytes
Desc: not available
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140807/d41578fa/attachment.jpg>

[Buildroot] [PATCH 1/1] wpa-supplicant: Add NL80211 support option

[Thomas Petazzoni]

Dear Jean-Baptiste Theou,

On Thu, 7 Aug 2014 13:52:33 -0700, Jean-Baptiste Theou wrote:

> IMHO, wpa_supplicant should provide the support of NL80211 by default 
> ('depends on BR2_PACKAGE_LIBNL) since it's the new standard.
> 
> I wasn't expected to have to select LIBNL manually.

There is a balance to find between:

 * Adding suboptions to package to enable support for optional features
   (what you did).

   Advantages : more obvious to the user, you can enable an optional
   feature in package A, but not in package B (like you can use OpenSSL
   support for your web server, but not necessarily for all other
   packages in your system)

   Drawbacks  : maintenance burden due to more Config.in options all
   over the place.

 * Make packages automatically enable optional features when the
   necessary dependencies are available.

   Advantages : less Config.in options to add everywhere.

   Drawbacks : less obvious to the user, less flexible (enabling
   OpenSSL makes it used in all packages that can optionally use it).

You could propose a patch that extends the Config.in help text of
wpa_supplicant to indicate this optional dependency, though we don't
have a specific policy about this.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] [PATCH 1/1] wpa-supplicant: Add NL80211 support option

[Jean-Baptiste Theou]

Hi Thomas.

I agree about finding the balance.

Regarding this specific matter, having an visual indication of the 
support or not for the new standard of the wireless support seems an 
important issue for me.

Having an message inside the help doesn't seems a good solution for me, 
since it's not a "side" feature of wpa_supplicant. I would like the 
LIBNL to be selected by wpa_supplicant, but I understand you are not 
agree with this option

I sent an V2 version of my patch, with an better indication of the link 
between libnl and wpa_supplicant.

If it doesn't work for you, I will send an patch with an indication on 
the help message, but IMHO, at least, this suboption is mandatory.

Best regards
On 08/08/2014 01:41 AM, Thomas Petazzoni wrote:
> Dear Jean-Baptiste Theou,
>
> On Thu, 7 Aug 2014 13:52:33 -0700, Jean-Baptiste Theou wrote:
>
>> IMHO, wpa_supplicant should provide the support of NL80211 by default
>> ('depends on BR2_PACKAGE_LIBNL) since it's the new standard.
>>
>> I wasn't expected to have to select LIBNL manually.
> There is a balance to find between:
>
>   * Adding suboptions to package to enable support for optional features
>     (what you did).
>
>     Advantages : more obvious to the user, you can enable an optional
>     feature in package A, but not in package B (like you can use OpenSSL
>     support for your web server, but not necessarily for all other
>     packages in your system)
>
>     Drawbacks  : maintenance burden due to more Config.in options all
>     over the place.
>
>   * Make packages automatically enable optional features when the
>     necessary dependencies are available.
>
>     Advantages : less Config.in options to add everywhere.
>
>     Drawbacks : less obvious to the user, less flexible (enabling
>     OpenSSL makes it used in all packages that can optionally use it).
>
> You could propose a patch that extends the Config.in help text of
> wpa_supplicant to indicate this optional dependency, though we don't
> have a specific policy about this.
>
> Best regards,
>
> Thomas

-- 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140808/e6ead6ef/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.jpg
Type: image/jpeg
Size: 8458 bytes
Desc: not available
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20140808/e6ead6ef/attachment.jpg>

[Buildroot] [PATCH 1/1] wpa-supplicant: Add NL80211 support option

[Thomas Petazzoni]

Dear Gustavo Zacarias,

On Thu, 07 Aug 2014 17:37:16 -0300, Gustavo Zacarias wrote:
> This is done automatically if the libnl package is selected/around.
> We really don't want to thread too thin into hostapd & wpa_supplicant
> config options.

Hum, so, Jean-Baptiste patch is adding CONFIG_DRIVER_NL80211 to the
configuration when libnl is available. But the current
wpa_supplicant.mk does (after removing the static lib special case)

ifeq ($(BR2_PACKAGE_LIBNL),y)
        WPA_SUPPLICANT_DEPENDENCIES += libnl
        WPA_SUPPLICANT_CONFIG_ENABLE += CONFIG_LIBNL32
else
        WPA_SUPPLICANT_CONFIG_DISABLE += CONFIG_DRIVER_NL80211
endif

So it enables CONFIG_LIBNL32 when libnl is used. Which is not exactly
the same thing. But interesting, in the else case, it's
CONFIG_DRIVER_NL80211 that gets disabled.

I'm a bit confused :)

Thanks,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com