On 04/28/2007 05:56 PM, Cody Jung wrote:

> There was a Vendor Id (ENS4081) and two Logical device Ids (ENS1011 and
> ENS2020).

Okay, that's a VIVO90. snd-sscape does not work for this card, and not just 
as a matter of a missing PnP ID. I suppose PnP Soundscapes exist for which 
the driver is correct but at least for the VIVO90 it fails the detect, tries 
to talk to the codec at a completely wrong address, and generally seems to 
not be applicable.

Attached is a minimal driver for the VIVO90. It only drives the AD1845 codec 
but chances are fairly good that's actually all you care about. Hardware 
(firmware rather) midi is what makes the Soundscape better/nicer than most 
but it's also the hard (undocumented) part and I haven't concerned myself 
with it for now.

The msleep(500) in there is not nice, but the card needs a delay after 
activation  until the codec is available and while I did look, I haven't 
found what to poll yet (it's not interrupting). The 500 ms hasn't failed to 
load for me yet, but it's not hugely roomy; 400 ms hasn't _succeeded_ for me 
yet so if it won't load for you please increase the 500 to 1000 as a first try.

This is not being submitted. It's only a very minimal thing and since Hannu 
said that the VIVO and the VIVO90 are in fact different models (mine does 
advertise itself as a "ENSONIQ SoundscapeVIVO" without the 90 in its PNP 
string) I'm not even sure if this is VIVO90 specific. The PNP ID probably 
is. It's only being posted in the hope that it'll work for your needs.

It'll probably grow into a real driver some day, but not soon; given that 
it's all very undocumented, getting more to work might be fairly painful. 
The old/other driver will help to some degree but I'm not sure to which.

The patch was generated against 2.6.21 and should apply without trouble to 
most recent kernels. Loading works just as "modprobe snd-vivo", no params.

Rene.