Am 2020-02-04 11:31, schrieb Michael Richardson: 

> Philipp Blum <[email protected]> wrote:
>> I just compiled OpenWRT with 6LoWPAN_bluetooth and configured it.
>> (https://imgur.com/a/8ctUHG9 & https://imgur.com/a/frYKx7N)
>> The interface gets an /60 prefix delegated (2a02:8aaa:baaa:b700::1/60),
>> which is great.
> 
> I assume that this is on the openwrt side of things.  I'm surprised it gives
> a /60 and not a /64, but I guess that is an openwrt setting.

Yes, I configured it that way to request a /60 PD. Because Linux doesn't
have a proper support for RPL, I don't want to have the /64 on the BT
interface. I can also connect more than one device to this interface.
Also a good reason to have /60 and not /64. 

> What is your RIOT-OS device?

The gnrc_networking example on the nrf52. 

> Which openwrt router is it that you have BT on? (Just curious)

Archer C7 v2 with a Fritzbox Cable in front of it in bridge mode
(Modem). I added a BLE USB stick to it and compiled the necessary
drivers and 6lowpan_bluetooth kernel modules for it. 

>> So now I want to configure RIOT OS to work with it.
> 
> I would think you need to do the L2 Bluetooth pairing first, but:
> 
>> I am able to ping the router from my node.
> 
> that would seem to suggest that this already works.
> You get replies?  Maybe you can explain exactly what you did that works.
> 
>> But the configuration of the interface doesn't work.

Yes, L2 BLE connection is there. I am able to ping the link local
address of the bt0 interface (OpenWRT router). So the connection is
there and they are able to reach each other. Now its (just) a question
about the correct network setup. PD to the bt0 interface looks also
fine.

 2020-02-04 12:15:13,389 # ble info 
2020-02-04 12:15:13,395 # Own Address: CC:C6:52:01:8E:8C ->
[FE80::CCC6:52FF:FE01:8E8C] 
2020-02-04 12:15:13,396 # Free slots: 2/3 
2020-02-04 12:15:13,397 # Advertising: no 
2020-02-04 12:15:13,399 # Connections: 1 
2020-02-04 12:15:13,404 # [ 0] 00:1A:7D:DA:71:13 (S) ->
[FE80::001A:7DFF:FEDA:7113] 
2020-02-04 12:15:13,405 # Contexts: 
2020-02-04 12:15:13,409 # [ 0] state: 0x0022 - GAP-slave L2CAP-server 
2020-02-04 12:15:13,411 # [ 1] state: 0x8000 - unused 
2020-02-04 12:15:13,414 # [ 2] state: 0x8000 - unused 
 > ping6 fe80::1a:7dff:feda:7113 
2020-02-04 12:27:11,497 # ping6 fe80::1a:7dff:feda:7113 
2020-02-04 12:27:11,591 # 12 bytes from fe80::1a:7dff:feda:7113:
icmp_seq=0 ttl=64 time=89.223 ms 
2020-02-04 12:27:12,591 # 12 bytes from fe80::1a:7dff:feda:7113:
icmp_seq=1 ttl=64 time=91.610 ms 
2020-02-04 12:27:13,591 # 12 bytes from fe80::1a:7dff:feda:7113:
icmp_seq=2 ttl=64 time=94.083 ms 
2020-02-04 12:27:13,592 # 
2020-02-04 12:27:13,595 # --- fe80::1a:7dff:feda:7113 PING statistics
--- 
2020-02-04 12:27:13,600 # 3 packets transmitted, 3 packets received, 0%
packet loss 
2020-02-04 12:27:13,605 # round-trip min/avg/max = 89.223/91.638/94.083
ms 

>> As I understand it, I should now be able to add it with
>> ifconfig 9 add 2a02:8aaa:baaa:b700::1/64
> 
> Do you mean ::2?
> I assume you are running this on the RIOT-OS side?

Yes, this is the RIOT side. Yes, ::2. 

> ifconfig 9 add 2aaa:8aaa:baaa:b700::2/64      

Which results in no response at all and the interface didn't change. It
only has the link local address. 

>> But this doesn't return anything. There is also the 6ctx I am not
>> familiar with.
> 
> It provides an context for a compression mechanism in 6lo.

Okay, this is not important at the moment, I guess. 

> You can't bridge, because the L2 headers are very different.
> You can route, using RPL, though, which is the recommended mechanism in IPv6.
> Going first hop to 6lowman as being BT has some interesting use cases, I 
> think.

Yes, I thought it should be similar to the SLIP configuration, but with
Bluetooth as interface. You should also be able to connect more than one
node to the bt0 interface. So that you create a star topology at the
router to many 6LoWPAN trees. 

  
_______________________________________________
users mailing list
[email protected]
https://lists.riot-os.org/mailman/listinfo/users

Reply via email to