Subject: Re: AC97 quirks?
To: Peter Seebach <>
From: None <>
List: current-users
Date: 07/05/2002 16:34:57
* Peter Seebach <> [020704 16:08]:
> In message <>, "Jare
> d D. McNeill" writes:
> >  enum ac97_host_flags {
> >          AC97_HOST_DONT_READ = 0x1,
> >          AC97_HOST_SWAPPED_CHANNELS = 0x2        /* l/r is reversed */
> >  };
> Hmm.  I guess I didn't see where those got set.
> >Are there any other quirks you have in mind?
> Whatever it is that makes the CT5880 with the "TriTech unknown" codec work.
> :)

I may be wrong, but I'm guessing you have a SB PCI128?

If so, this works in FreeBSD; I switched this box to NetBSD and luckily
have an old dmesg lying about.

The corresponding driver in FreeBSD is pcm(4), defined in:


And so I can tell you what that TriTech codec is:

>> NetBSD version:

1rasputin@idoru:eap-hack$ dmesg|grep eap
eap0 at pci0 dev 10 function 0: Ensoniq CT5880 CT5880C (rev. 0x02)
eap0: interrupting at irq 9
eap0: TriTech unknown codec; no 3D stereo
audio0 at eap0: full duplex, mmap, independent
midi0 at eap0: AudioPCI MIDI UART

>> FreeBSD (4.5) version:

1rasputin@idoru:eap-hack$ grep pcm shikima.dmesg
pcm0: <Creative CT5880-C> port 0xf000-0xf03f irq 9 at device 9.0 on pci0
pcm0: ac97 codec id 0x54524123 (TriTech TR28602)
pcm0: ac97 codec features 5 bit master volume, no 3D Stereo Enhancement
pcm: setmap f8bc000, 1000; 0xc11c8000 -> f8bc000
pcm: setmap f87e000, 1000; 0xc11ca000 -> f87e000

I am dumb enough not to realise how difficult it is to add support for this,
so was going to try merging the relevant entries from their ac97.c into

Rasputin :: Jack of All Trades - Master of Nuns