> > There must be something wrong with this. Essentially, the sound
> > configuration or the makefile appear broken: with certain configurations
> > the "ad1848" files would show up multiple times, which cannot be what this
> > should be all about.
>
> The makefile is still fairly broken. I need to start throwing some def_*
> stuff into the config and taking things out of the makefile
Hmm.. not much has changed in terms of dependencies, as far as I can tell.
dev_table.h will make sound.o depend on the chipset-specific drivers which
in turn depend on sound.o. Cyclic dependencies means the modules will
never load.
I'm thinking of a slighly different approach, something similar to the
scsi subsystem. More precisely, a sound.o which does not depend on
anything and provides high-level functions only, so it can be loaded
separately without causing any low-level driver to be loaded. It would
also provide some sort of registration functions, so low-level drivers can
register themselves for the specific devices they provide - dsp, midi,
mixer. sound.o will also call request_module with a generic name (aliased
to a specific driver in modules.conf) whenever a device is opened and a
driver is not attached to it.
As far as I can tell, the infrastructure for this is already in place -
nice work Hannu et al - all it needs is the glue functions to keep it in
place. Does it sound reasonable?
[FWIW, I'm running a minimally modified 2.0.33 here with a SB16, a
SB-AWE32 and a CS4232, all of which are recognized and usable - at the
same time :-)]
Ionut
--
It is better to keep your mouth shut and be thought a fool,
than to open it and remove all doubt.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu