diff mbox

[RFC] mesh: Disable OBSS scan for mesh

Message ID 1469504898-6244-1-git-send-email-masashi.honma@gmail.com
State RFC
Headers show

Commit Message

Masashi Honma July 26, 2016, 3:48 a.m. UTC
IEEE802.11 2012 "10.15.3 Channel selection methods for 20/40 MHz
operation" looks only mentions about AP and DO STA. So disable OBSS scan
for mesh.

Signed-off-by: Masashi Honma <masashi.honma@gmail.com>
---
 wpa_supplicant/wpa_supplicant.c | 2 ++
 1 file changed, 2 insertions(+)

Comments

Masashi Honma Nov. 28, 2016, 12:27 a.m. UTC | #1
On 2016年07月26日 12:48, Masashi Honma wrote:
> IEEE802.11 2012 "10.15.3 Channel selection methods for 20/40 MHz
> operation" looks only mentions about AP and DO STA. So disable OBSS scan
> for mesh.

Is there any comment for this ?

Masashi Honma.
Jouni Malinen Nov. 29, 2016, 1:41 p.m. UTC | #2
On Tue, Jul 26, 2016 at 12:48:18PM +0900, Masashi Honma wrote:
> IEEE802.11 2012 "10.15.3 Channel selection methods for 20/40 MHz
> operation" looks only mentions about AP and DO STA. So disable OBSS scan
> for mesh.

While IEEE Std 802.11s-2011 may not have addressed this area (nor IEEE
Std 802.11-2012 or the upcoming -2016 version), I'm not sure this is
really the correct thing to do.. The DO STA concept sounds something
similar to what a mesh BSS should likely have anyway.. Or alternatively,
every single mesh STA would be required to do similar operations.

As far as OBSS scan is concerned, it sounds fine to skip this for the
case when joining an already operating mesh BSS. However, when starting
up a new mesh BSS, I don't see why a mesh STA would have more freedom in
skipping 20/40 MHz co-ex requirements when compared to a STA that is
starting an IBSS or an AP starting an infrastructure BSS.
Masashi Honma Dec. 1, 2016, 10:52 p.m. UTC | #3
On 2016年11月29日 22:41, Jouni Malinen wrote:
> While IEEE Std 802.11s-2011 may not have addressed this area (nor IEEE
> Std 802.11-2012 or the upcoming -2016 version), I'm not sure this is
> really the correct thing to do.. The DO STA concept sounds something
> similar to what a mesh BSS should likely have anyway.. Or alternatively,
> every single mesh STA would be required to do similar operations.
>
> As far as OBSS scan is concerned, it sounds fine to skip this for the
> case when joining an already operating mesh BSS. However, when starting
> up a new mesh BSS, I don't see why a mesh STA would have more freedom in
> skipping 20/40 MHz co-ex requirements when compared to a STA that is
> starting an IBSS or an AP starting an infrastructure BSS.

Thank you for comment. I see. I will drop this patch.

Masashi Honma.
diff mbox

Patch

diff --git a/wpa_supplicant/wpa_supplicant.c b/wpa_supplicant/wpa_supplicant.c
index 7ed7efa..25ff976 100644
--- a/wpa_supplicant/wpa_supplicant.c
+++ b/wpa_supplicant/wpa_supplicant.c
@@ -1848,6 +1848,8 @@  void ibss_mesh_setup_freq(struct wpa_supplicant *wpa_s,
 			break;
 		}
 	}
+	if (ssid->mode == WPAS_MODE_MESH)
+		obss_scan = 0;
 
 	/* For IBSS check HT_IBSS flag */
 	if (ssid->mode == WPAS_MODE_IBSS &&