Review of the rfkill drivers in ubuntu

[Stefan Bader]

Tim Gardner wrote:
> Stefan Bader wrote:
>> As part of the effort to clean up the ubuntu directory I had a look at 
>> the rfkill patches there (as those were deemed needing to go into 
>> staging soon).
>>
>> However I am not sure we really want this (how bad can a driver be 
>> before one gets clubbed with it?).
>>
>> * fsam7400 is based on pbe5 which is based on the av5100.
>> * all three most be loaded manually and will take the state of the radio
>>    from the argument (not from the current hw state)
>> * Only fsam7400 makes some effort to check the bios signature. The other
>>    two just punch on the ports.
>> * The control interface is procfs. No use of any rfkill framework.
>> * fsam7400 and pbe5 state the port information has been extracted by
>>    disassembling windows drvers (maybe the smallest issue)
>>
>> I got the feeling that in the current state the drivers are not ready for
>> staging and to make them better we would need to find someone with 
>> that HW.
>> Maybe we should disable them and see who shouts? Thoughts?
>>
>>
> 
> Perhaps we should send email to the developers of the various drivers in 
> order to gauge their interest. If we don't get a positive response, 
> _then_ we can disable the driver.
> 

If we can get feedback there, this would be better. This definitely requires hw 
to test with. I constantly update the spec with findings. There seem to be two 
sources and one of them even claims to have a user-space program doing the 
same. So I will try to get into contact with the author. Generally, with the 
code being so far away from anything using the rfkill framework, this eases a 
bit the requirement to act quick.

Stefan
> rtg


-- 

When all other means of communication fail, try words!