[rt2x00-users] rt2800pci SoC

[Benoit PAPILLAULT]

Gertjan van Wingerde a écrit :
> On Mon, Mar 8, 2010 at 8:34 AM, Helmut Schaa
> <helmut.schaa at googlemail.com> wrote:
>   
>> Hi,
>>
>> as I wrote last week already I am at least able now to get rt2800pci loaded
>> on the rt3050 SoC. However, just triggering a scan (on the default STA
>> interface) reveals no results.
>>
>> Looking at the TX path, it seems mac80211 tries to send 11 probe requests
>> (11 enabled channels) and in response I also get some interrupts back from
>> the device. The rt2800pci_interrupt routinge gets called for every tx frame.
>> Unfortunately I don't get TX_FIFO_STATUS interrupts but AC0_DMA_DONE.
>>     
I think it might be because TX descriptors are not correctly 
initialized. See :
http://git.popipo.fr/?p=rt2x00.git;a=commitdiff;h=faddaa8ab3aaa86e6c7fea53463b7a15fa09ac57
>> Sniffing the wireless traffic does _not_ reveal any probe requests send by
>> that device.
>>
>> On the other hand I don't get any rx traffic (RX_DONE interrupts), even in
>> monitor mode, but I didn't dig into that so far.
>>
>> Any suggestions/ideas on how to debug that further? Does rt2800pci show
>> similar symptoms on PCI devices as well or just on SoC?
>>
>>     
>
> This seems to match other reports that we've had on rt2800pci. On the
> other hand, I've seen others claiming to be sucessful with the current
> set-up. So far, I've never been successful connecting my rt2800pci to
> an actual network, although I could scan on that machine.
>
> I've got it on my TODO-list to audit our interrupt handling against
> what the Ralink vendor driver does, but I've lacked the time for it
> lately. :-(
>
> ---
> Gertjan
>   
Same for me :-)

Regards,
Benoit