Re: [Rt2400-devel] [PATCH] rt2x00: Fix NULL pointer error in adhoc/master mode

[Ivo van Doorn <ivdoorn@gmail.com>]

On Wednesday 09 July 2008, Bryan Batten wrote:
> Ivo van Doorn wrote:
> > On Saturday 05 July 2008, Bryan Batten wrote:
> >> Ivo van Doorn wrote:
> ...
> >>> Note that the "raise beacondone interrupts without a beacon" is
> >>>  also a bug, and will be addressed later.
> ...
> >> FWIW, I've never been able to figure out a way to shut it off in 
> >> the legacy drivers. Unless there's some mechanism I've missed, I 
> >> would consider this a bug, but its a hardware bug because - in 
> >> the rt2500 driver at least - even disabling it in the interrupt 
> >> mask register has no effect.
> 
> Looking thru the RT2400 and RT2500 datasheets, I notice that on pp. 12
> of the 2400 datasheet, CSR8 bits 0 and 1 are described. On the same
> page of the RT2500 datasheet, CSR8 bits 0 and 1 are not described. If
> that is not a typo, that means to me that on the RT2400 you can
> mask off beacon timer and wakeup timer interrupts with CSR8, but on
> the RT2500 you cannot.

Yes, but if you look again you see that the legacy driver does define it,
and even includes the description of bits 9 -> 19
And the legacy driver source code has always been more reliable then
the datasheet, so I would put my money on the code. ;)

> Would it be possible to verify with Ralink that this is not a typo in
> the RT2500 datasheet?

I'll see what I can find out.

Ivo