[CFT]RT28xx/RT30xx wireless was [CFR]RT305xF support, w/o attachment

[Aleksandr Rybalko]

On Thu, 31 Mar 2011 23:44:03 -0700 (PDT)
PseudoCylon <moonlightakkiy at yahoo.ca> wrote:

>> > > rt28600: <Ralink  RT2790 PCIe> mem 0xf7f00000-0xf7f0ffff irq 17 at
>> > > device 0.0 on  pci3
>> > > rt28600: invalid EEPROM LNA gain #2: 0x00
>> > > rt28600: invalid  EEPROM LNA gain #3: 0x00
>> > > rt28600: invalid EEPROM powersave level
>> > >  rt28600: MAC/BBP RT2860 (rev 0x28720200), RF RT3022 2.4G 2T2R
>> > Wow, your  device have same revision 0x28720200 like embedded into
>> > RT3052F system on  chip.
>> > So now I understand, why driver won't work with your card.
>> > I  previously expect that this id related only for SoC version, but SoC
>> > version  don't have many things that PCI version have (MCU, EEPROM,
>> > etc.)
>> > 
>> 
>> Hi,
>> 
>> I have 0x28720200 calling rt2872_rf_set_chan() instead of rt2860_rf_set_chan().
>> And, 0x2000 for initial RT2860_REG_MAX_LEN value in rt2860_init_locked().
>> 
>> If these are correct I can patch the driver up.

Basic problem here is a usage external EEPROM data(instead read from real EEPROM) and direct BBP control(instead use MCU).
So currently prefer to use Alexander Egorenkov version, until we done porting OpenBSD one and join it with ral(4) code.

>> 
>> AK
>> 


-- 
Alexandr Rybalko <ray at dlink.ua> 
aka Alex RAY <ray at ddteam.net>