Clemens Ladisch wrote:
Ray Lambert wrote:
Takashi Iwai wrote:
Since snd-mpu401 is a different driver from snd-intel8x0, it's
registered as an indivial instance, the secondary card, as a result.
Okay...  I guess I'm just not following this "primary" and "secondary
card" thing.  Do you mean to say that the driver (snd-mpu401) is a
/secondary driver/ for the card (snd-intel8x0 being the primary)?  Or that
the device (mpu401) is a /secondary device/ on the integrated sound card?
The AC'97 controller and the MIDI port are two completely different and
independent devices (one a PCI device in the south bridge, the other an
ISA device in the SuperI/O chip) that just happen to be soldered on the
same mainboard.

Thanks for the explanation.  It makes even more sense now.  I guess I'm just used to thinking of the game/midi port as being part of the sound card (this is my first experience with integrated audio, in case you couldn't tell...  ;)  ).

In any case, I'm still trying to get this working.  I still need confirmation of the correct way to create the midi device node for OSS raw midi (i.e. for sysex dumps).  I'm currently leaning towards: "mknod /dev/midi c 14 18".

I'm also looking for ways to determine whether or not the hw is working correctly (independent of the device node).  "cat /proc/asound/card1/midi0" looks promising as it has a Rx count (which will hopefully show me that bytes are arriving from my midi device).  But I'm not sure if it increments without a proper device node and someone reading from it.

I wrote a test program that opens and reads from the device but I get zero bytes no matter what I've tried.

If you (or anyone) has any advice it would be greatly appreciated.

Thanks for your help!