[I Cc: this message to David Hinds, and Thomas Hood, because
it deals with PnpBios support, taken from pcmcia-cs. FYI the beginning
of this thread is at http://www.geocrawler.com/lists/3/SourceForge/12349/0/]
Hello Uros and ALSA hackers,
On Sat, Dec 22, 2001 at 11:11:03PM +0100, Uros Bizjak wrote:
> Hello Fabrice!
>
> I can confirm, that OPL3 part of CS4236B chip works with current CVS. I have
> produced FM noises with playmidi -f xxx.mid and playmidi -4 xxx.mid commands.
> All I have to do was to unmute and raise FM volume and unmute and raise volume
> on Digital Mixer with alsamixer. However, I didn't test pmidi (and sbiload),
> but I don't
> expect any problems with these.
> I would suggest you to play a little with CS4239 extended registers. There
> should be
> a bit, which turns on FM synth or enables route for FM data to A/D converters.
> I would like to help you more, but it is very hard to 'debug' a problem without
To my surprise, the internal FM synthesizer is now working, after a very
recent reboot of my laptop. The configuration part that changed since
this reboot is that I no longer use the 'setpnp' command to rewrite
the PnP ressources of my laptop (Thinkpad 770Z, cs4239 audio chip).
I also upgraded my kernel : 2.4.17-pre6 --> 2.4.17-final, and
pcmcia-cs.30.Nov.2001 --> pcmcia-cs.20.Dec.2001.
PnPbios support and lspnp/setpnp commands are part of the pcmcia-cs
package. The access to PnP resources is a bit confusing, because a
flag '-b' allows to modify a copy of the resources, that'll
be in place in the next reboot. And without this flag, you modify actual
PnP resources. So at each time, you have to handle two copies of the
Pnp resources.
For some strange reason, the '-b' flag provides the opposite behaviour
than expected. After a fresh reboot, all resources marked for next
reboot are disabled :
[bellet@bonobo bellet]% lspnp -v -b 0e 0f 10 11
0e CSC0100 multimedia controller: audio
io disabled
io disabled
io disabled
irq disabled
dma disabled
dma disabled
0f CSC0110 multimedia controller: audio
io disabled
10 CSC0101 multimedia controller: audio
io disabled
11 CSC0103 multimedia controller: audio
io disabled
irq disabled
[bellet@bonobo bellet]% lspnp -v 0e 0f 10 11
0e CSC0100 multimedia controller: audio
io 0x0530-0x0537
io 0x0388-0x038b
io 0x0220-0x0233
irq 9
dma 1
dma 0
0f CSC0110 multimedia controller: audio
io 0x0538-0x053f
10 CSC0101 multimedia controller: audio
io 0x0200-0x0207
11 CSC0103 multimedia controller: audio
io 0x0330-0x0333
irq 10
Then, I issue the necessary commands to reenable resources :
for arg in "-b"
do
setpnp $arg 0e io 0x0530-0x0537 io 0x0388-0x038b io 0x0220-0x0233 irq 9 dma 1
dma 0
setpnp $arg 0f io 0x0538-0x053f
setpnp $arg 10 io 0x0200-0x0207
setpnp $arg 11 io 0x0330-0x0333 irq 10
done
With this new configuration, the ALSA driver is working fine,
except the internal FM, that remains muted. Moreover, this
little script was essential. Without it, modprobe
oopsed in loading the ALSA driver. I was running ALSA with this
configuration, when I asked you guys about my non working
midi/synth/fm support.
Yesterday, I upgraded to kernel-2.4.17, and pcmcia-cs-20.Dec.2001,
and I noticed that 'setpnp' is no longer required before starting
alsasound. With the ressources, as listed above, the ALSA drivers
now load fine, and provide all their expected capabilities, FM
included. Fine!
The strange remaining behaviour is: All PnP ressources are marked
disabled in the '-b' space at boot time. If I 'setpnp -b' audio ressources,
and restart alsasound, I loose FM capability. If I disable resources
again, the modprobe Ooopses in the alsasound script.
I'd be happy to process to further testing, if needed. Please just
tell me how. I'd like also to thank you ALSA guys for your precious
help and time. Keep on doing your great work.
--
fabrice
_______________________________________________
Alsa-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/alsa-devel
