No DMA RX on some BCM4321, on BCM43224 and BCM43225

["Rafał Miłecki" <zajec5@gmail.com>]

W dniu 19 lipca 2011 00:39 u?ytkownik Rafa? Mi?ecki <zajec5@gmail.com> napisa?:
> W dniu 18 lipca 2011 22:41 u?ytkownik Michael B?sch <m@bues.ch> napisa?:
>> On Mon, 18 Jul 2011 22:14:50 +0200
>> Rafa? Mi?ecki <zajec5@gmail.com> wrote:
>>
>>> When using DMA on one of my BCM4321s and BCM43224 or BCM43225 I don't
>>> get a single RX IRQ.
>>>
>>> Doing scanning requests sends TX packets to the hardware, and the
>>> ucode generates IRQs for confirming transmitted packets. RX does not
>>> seem to work however. I expected to see "native" beacons, or response
>>> to the scanning request.
>>>
>>> Do you have idea what may be wrong, where should I look at?
>>
>> Is wl still using the same dma engine for
>> RX as older versions did? (what was it? the first?)
>
> Yes. That's what wl does:
>
> 1) Disabling DMA engines (RX part)
> write32 0xfaafc220 <- 0x00000000
> ?read32 0xfaafc230 -> 0x00000000
>
> 2) Enabling DMA engines (RX part)
> ?read32 0xfaafc220 -> 0x00000040
> write32 0xfaafc220 <- 0x0000084d CTL: [38 << 1 (OFFSET) | 0x1 (EN) |
> 0x800 (PARITY)]
> write32 0xfaafc228 <- 0x18948000 LO: [lo addr & ~routing_bits]
> write32 0xfaafc22c <- 0x80000000 HI: [routing & ~ROUT_MASK | routing]

As you may now, wl does a lot of weird operations. Before it really
starts working, it resets everything 10 times.

For DMA it looks like this:
1) Check if DMA is 30 or 32 bit (by setting and reading back "Address
Extension"). wl does it even when TM State High clearly says 64 bit
DMA is availale (and wl knows it!):
ssb_read32(0xf9c) -> 0x500c0000
write32 0xfaafc200 <- 0x00030000
 read32 0xfaafc200 -> 0x00030000
2) Disable every DMA engine (SUSPEND for TX, zero ctl register for TX)
3) Disable every DMA engine (SUSPEND for TX, zero ctl register for TX) (yeah...)
4) Setup rings (pass routing, addresses, set enable, parity, RX header
length for RX)

Maybe disabling engines before setting them up is needed?

-- 
Rafa?