Interesting 14e4:4321

[Rafał Miłecki]

W dniu 6 maja 2011 23:36 użytkownik Hauke Mehrtens <hauke at hauke-m.de> napisał:
> On 05/06/2011 10:40 PM, Rafał Miłecki wrote:
>> Swen has found 14e4:4321 in his WRT160N version 1.1. This is quite
>> rare and interesting so I wanted to post all info about it :)
>>
>> I'm attaching lspci and dmesg.
>>
>> I find it interesting because it seems to be SSB SOC with standard
>> cores: CC, FastEthernet, USB, MIPS, etc. Plus there is PCI bridge with
>> 14e4:4321 card connected, which is SSB-based one with (the most
>> interesting) 80211.
>>
>> Unfortunately this is one another BCM4321 with DMA not working. I
>> really need to focus on this after fixing LP-PHY and adding BCMA.
>>
> Hi Rafał,
>
> RealOpty reported the same DMA error in the OpenWrt channel and it is
> nice that you are working on this. He was using a Netgear WNR834B V2.
> Here is his dmesg output: http://paste.ubuntu.com/599979/

It's nice to see my ssb check for 0x0F90 is working :)


> We already fixed some of the problems with these devices but have not
> investigated into the DMA errors.

Fortunately some Polish user sent me his already-replaced BCM4321
card. So I can reproduce DMA problems locally.


> The cardbus attribute in the nvram seams to be set wrong for some
> devices we added a workaround for it for the Netgear WNR834B V2. [0] It
> would be nice to get the nvram content ( nvram show ) to also detect
> this device and activate the workaround for it.

Swen: could you provide "nvram show" from your device for Hauke, please?


> The SPROM is not stored on the pci card but in the nvram on the flash
> chip of the device. This is now correctly read out in OpenWrt and we
> will send some patches to mainline soon. [1] [2]

Swen already updated his OpenWRT trunk and he got SPROM fallback working :)

-- 
Rafał