Hi,
I'm trying the get a Redpine RS9116 module working with networkmanager.
I've tried this on 5.3, 5.5 and next-20200128. I'm using the Redpine 1.5
"rs9116_wlan_bt_classic.rps" firmware.
If I configure the interface using iw, wpa_supplicant and dhclient all
works as expected.
If I try to configure the interface using nmtui most of the time no APs
show up to associate to. "iw dev wlan0 list" shows all of the APs in the
vicinity.
If I do manage to get an AP to show when I try to "Activate a
connection" I get the error below
Could not activate connection:
Activation failed: No reason given
I suspect this is a driver bug rather than a NM bug as I saw similar
issues with an earlier Redpine proprietary driver that was fixed by
updating that driver. What rsi_dbg zone will help debug this ?
Thanks
Angus
On Thu, 2020-01-30 at 10:18 -0800, Angus Ainslie wrote:
> Hi,
>
> I'm trying the get a Redpine RS9116 module working with
> networkmanager.
> I've tried this on 5.3, 5.5 and next-20200128. I'm using the Redpine
> 1.5
> "rs9116_wlan_bt_classic.rps" firmware.
>
> If I configure the interface using iw, wpa_supplicant and dhclient
> all
> works as expected.
>
> If I try to configure the interface using nmtui most of the time no
> APs
> show up to associate to. "iw dev wlan0 list" shows all of the APs in
> the
> vicinity.
>
> If I do manage to get an AP to show when I try to "Activate a
> connection" I get the error below
>
> Could not activate connection:
> Activation failed: No reason given
>
> I suspect this is a driver bug rather than a NM bug as I saw similar
> issues with an earlier Redpine proprietary driver that was fixed by
> updating that driver. What rsi_dbg zone will help debug this ?
NM just uses wpa_supplicant underneath, so if you can get supplicant
debug logs showing the failure, that would help. But perhaps the driver
has a problem with scan MAC randomization that NM can be configured to
do by default; that's been an issue with proprietary and out-of-tree
drivers in the past. Just a thought.
https://blog.muench-johannes.de/networkmanager-disable-mac-randomization-314
Dan
Hi Dan,
On 2020-01-30 10:39, Dan Williams wrote:
> On Thu, 2020-01-30 at 10:18 -0800, Angus Ainslie wrote:
>>
>> I suspect this is a driver bug rather than a NM bug as I saw similar
>> issues with an earlier Redpine proprietary driver that was fixed by
>> updating that driver. What rsi_dbg zone will help debug this ?
>
> NM just uses wpa_supplicant underneath, so if you can get supplicant
> debug logs showing the failure, that would help. But perhaps the driver
> has a problem with scan MAC randomization that NM can be configured to
> do by default; that's been an issue with proprietary and out-of-tree
> drivers in the past. Just a thought.
>
> https://blog.muench-johannes.de/networkmanager-disable-mac-randomization-314
>
Thanks that was the fix.
Angus
> Dan
On 2020-01-30 11:08, Angus Ainslie wrote:
> Hi Dan,
>
> On 2020-01-30 10:39, Dan Williams wrote:
>> On Thu, 2020-01-30 at 10:18 -0800, Angus Ainslie wrote:
>>>
>>> I suspect this is a driver bug rather than a NM bug as I saw similar
>>> issues with an earlier Redpine proprietary driver that was fixed by
>>> updating that driver. What rsi_dbg zone will help debug this ?
>>
>> NM just uses wpa_supplicant underneath, so if you can get supplicant
>> debug logs showing the failure, that would help. But perhaps the
>> driver
>> has a problem with scan MAC randomization that NM can be configured to
>> do by default; that's been an issue with proprietary and out-of-tree
>> drivers in the past. Just a thought.
>>
>> https://blog.muench-johannes.de/networkmanager-disable-mac-randomization-314
>>
>
Any chance there's something similar for bluetooth ? hcitool shows the
device and can scan but bluetoothctl can't find the device.
$ hcitool dev
Devices:
hci0 88:DA:1A:9E:XX:XX
[bluetooth]# list
[bluetooth]# devices
No default controller available
Thanks
Angus
>
>> Dan