dmasound hangs on new G4s (2)

[Iain Sandoe]

Mon, Mar 26, 2001, Tom Rini wrote:
> On Mon, Mar 26, 2001 at 01:00:22PM +0100, Iain Sandoe wrote:
>>
>> >>Okay, dmasound compiled in hangs the new G4s, but this change had no effect
>> >>on it otherwise (I also have a few Bogus interrupts, w/ or w/o this).
>> >
>> > One of those new G4s that have the new "digital sound" chip ?
>> >If I remember correctly, they have no AWACS, but an ibook-like chip on the
>> > I2S bus. I beleive it's normal dmasound doesn't work properly with it.
>>
>> "Tumbler" audio (whatever that is ;-))) [WWDC, Ben?]
>>
>> OK - could you send me the device-tree stuff? - and I'll compare it with
>> what we've got.
>
> It's whatever is in the new G4s.  And I do believe there was an i2s bus in
> OF when I was poking around before.
>
>> http://tibit.org/ppc/ has a nice (if somewhat slow) script that turns the
>> dev-tree into html.
>
> I'll shoot this off to you privately in a bit.

thanks.
This confirms it is Tumbler - and also that the new machines have a
higher-than-44k1 output sample rate (48k)... but no audio inputs :-(

>> With the BK _2_4 tree (and no other patches) i2s-a is checked and connected
>> for output - *but* there may be a 'mixer-related' issue (because that's
>> different too).  I've fixed that up for the iBook (so that it doesn't try
>> and do any AWACS-related stuff).... but that won't work for the new G4
>> necessarily.
>
> Okay.  I'll see what happens when I enable modules in a bit as well...

I think it'll break with (-ENODEV) until someone patches dmasound_init() to
recognise the tumbler... but ... I suppose it shouldn't hang... I'll try to
get to it RSN.

Iain.

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/