Re: CMI8338/8738 PCI sound chip - driver flaws, I believe

From: Xiong Jiang (jxiong@gwu.edu)
Date: Sun Feb 13 2000 - 02:22:48 EST


Thanks for all the responses, especially this one. Though I am
reluctant to try this 'commercial' OSS, encouraged by this response,
I've tried and found out:

Yes, OSS/Commercial does support it, completely, including OPL3,UART401,
mic gain control, and radio-box style recording source selection, all
of which could not be found out when using our 'official' OSS/free
cmpci.o
driver. Moreover, when I tried recording with OSS/Commercial modules
(also cmpci.o, and some others), removed them out of the memory, and
loaded our own cmpci.o, this time, recording STILL work. After a total
reboot of the mechine, recording fails again. This leads me to think
that there are flaws (minor or not) in current cmpci.c, that whether
it is intentionally not implemented in full or not is unknown.
(Could anybody further test this ?)

However, it can be expected the improvement task is minor, given
somebody more familiar to sound card interfaces and kernel modules
than me. And detailed related hardware interface could be revealed
by reverse engineering if vendor document is unavailable.

This is cheap hardware, but cheap hardware is also a hardware
and should have the right to work as it is designed. I urge if somebody
could put his hand on this or teach me some on sound card kernel modules
or help me tackle it in other ways.

Thanks,

iafilius@xs4all.nl wrote:
>
> Hello Xiong,
>
> Some questions about it:
>
> When doing a 'cat /dev/audio > aap' produce an output file size bigger
> then zero ?
> Shows 'cat /proc/interrupts' activity on soundcard IRQ('s) ?
> What shows gom (the console mixer) ?
>
> if everything looks OK:
> Is this soundcard supported by the oss souddriver (commercial, but demo
> available), if yes did you try this one ? (http://www.opensound.com/)
> Is this soundcard yet supported by the current 2.2.xx or 2.3.xx
> kernels? did you try these ?
>
> Greatings,
>
> Arjan Filius

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Tue Feb 15 2000 - 21:00:24 EST