Re: USB keyboard not working

Alexander Graf <[email protected]> Mon, 19 Apr 2010 23:53:22 +0200
Newsgroups gmane.linux.ports.game-cube.devel
Message-ID <[email protected]>
On 19.04.2010, at 21:35, Albert Herranz wrote:

>> So the next thing I 
>> stumbled over is ehci support. As I said earlier I merged your gc-linux changes 
>> into kvm.git which is pretty close to tip. Apparently I'm getting a DSI in the 
>> interrupt handling code now.
>> 
> 
> Obviously your fault :-P

Obviously :). I'd actually guess at interface changes between 2.6.32 and 2.6.34.

> 
>> Basically there are two questions arising 
>> from this:
>> 
>> 1) Is there an easy way to debug such things? I'm having a 
>> hard time reading from a TV screen, the contents scroll out, copy&paste 
>> doesn't work, etc. Netconsole isn't an option I guess because wifi doesn't play 
>> well with it. So what is left to get dumps?
>> 
> 
> The easiest way to debug stuff is to use an "usbgecko".
> It's an adapter that plugs into a memory card slot of the gamecube/wii on one side and a USB port on your PC on the other side. It provides a USB serial connection on the PC and a simple SPI interface on the gamecube/wii which allows to send/receive data to the PC.

Phew - sounds like me "need to buy" list is getting bigger and bigger :).

> 
>> 2) Would you mind rebasing 
>> your master to something more current 
>> :-).
> 
> 
> I decided to mainline code instead of continuing development out of tree.
> See www.gc-linux.org/wiki/Mainline_Kernel for a current status of what's already in.

Yeah, but that feature set is certainly not enough to actually do something with the Wii :-).

> But, as mainlining is taking more time than I expected (for example, I'm stuck waiting for some USB and swiotlb core changes that the EHCI driver patches for mainline depends on) I may eventually integrate as-is the old MIKE series into a mainline-based branch. While doing that, I'll get rid of all IOS-based support which I don't plan to support on the mainline kernel.

Makes sense. I was merely stating that I'd love to have your "master" tree be synced to what's upstream now. There are a lot of conflicts between your tree that actually has all the drivers and tip.


Btw - I have more odd USB results:

I figured I'd move my root fs from the SD card to a USB drive to speed up things a bit. Hence the try with EHCI. So I went back to OHCI and can't seem to get any USB HDs working as is. As soon as the partition table should be read out I run into a timeout and the HCI driver just sends a reset after n seconds.

The fun thing is that as soon as I also plug in my USB keyboard, everything works fine in OHCI mode. I can use the hard drive just fine. This is all on the MINI based system.

Can you reproduce this on your system too?


Alex


------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev