Re: sound problems persist in 2.3.99pre3

From: Bob_Tracy (rct@gherkin.sa.wlk.com)
Date: Wed Mar 22 2000 - 08:40:15 EST


Thomas Molina wrote:
> I've made progress on my sound problems. I can confirm that in order to
> use the OPL3 synthesizer on my PAS16 the PAS16 has to be initialized
> BEFORE opl3 is initialized. If pas16, soundblaster, and opl3 are built
> into the kernel, the opl3 cannot be initialized correctly. I followed
> this by reconfiguring the pas16 and soundblaster as builtins and the
> opl3 as a module. when I did an insmod opl3 io=0x388 it worked
> correctly and playmidi did its magic properly.

What you said... In my case, I don't use the soundblaster side of
the card (last four pas2 parameters 0,-1,-1,-1), but building the pas16
and opl3 into the kernel still results in the opl3 not working because
the opl3 gets initialized before the pas16.

> Alan, do you agree with my analysis? I can possibly create a patch to
> have the pas2 code to do the proper opl3 initialization, but it will
> take some time for me to get up to speed on the new way.

Back on March 10 I mentioned to Alan that there was no way to control
the order in which built-in sound devices were initialized, and this
meant we were pretty much forced to go with modules. He indicated he
wasn't working on 2.3.X stuff and hadn't been for a while, so for what
it's worth, *I* agree with your analysis. I'm not sure what the
officially blessed method of manipulating built-in device initialization
order is, but my initial knee-jerk thought is to bypass opl3
initialization if CONFIG_SOUND_PAS is defined (compile-time directive),
and have the PAS16 init code call the opl3 init code if
CONFIG_SOUND_YM3812 is defined. Yeah, it's ugly, but we shouldn't be
forced to build device drivers for non-removable devices as modules,
and auto-probing for sound devices is a minefield I don't think we're
ready to walk through :-).

-- 
Bob Tracy               |  "C++ is to C as Lung Cancer is to Lung."
rct@wlk.com             |  --Anonymous Coward on slashdot.org

- 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 : Thu Mar 23 2000 - 21:00:36 EST