Skip to contentSkip to conversation

BUG: Yamaha OPL3SA2 does not work with ALSA on 2.6 kernels.

DISCUSSION Andrew Haninger 3 participants 6 messages Jul 25, 2005 – Jul 26, 2005

lore
Share & tools
lore
#0 lore raw

Hello.

I have a 5 year old Gateway Solo 2500 that is currently running Linux 2.6.12.2. If I install ALSA and try to have alsaconf bruteforce-detect the OPL3SA2 sound card, it will say that it has detected it, but loading the modules will fail. If I install Linux 2.4 and recompile/rerun alsaconf, the detection works fine and the card works. Copying the configuration detected under 2.4 into a modprobe.conf on 2.6 allows me to use the card in 2.6 with occasional crashes (which might be due to suspend2).

Searching around the net, I find many other people having trouble with these cards and the ALSA-Linux2.6 combination. On one page, someone suggested that there were changes made between 2.4 and 2.6 to the ISA code that broke ALSA's detection routines.

I'm not sure what information might be needed in order to get this card working well once and for all, but if someone will let me know, I'd be happy to provide.

Thanks.

-Andy

#1 lore raw
Quoted 1 line · Andrew Haninger

Hello.

Hi Andrew,

Quoted 8 lines

I have a 5 year old Gateway Solo 2500 that is currently running Linux 2.6.12.2. If I install ALSA and try to have alsaconf bruteforce-detect the OPL3SA2 sound card, it will say that it has detected it, but loading the modules will fail. If I install Linux 2.4 and recompile/rerun alsaconf, the detection works fine and the card works. Copying the configuration detected under 2.4 into a modprobe.conf on 2.6 allows me to use the card in 2.6 with occasional crashes (which might be due to suspend2).

the ALSA people might be able to help you (Cc'ed).

Quoted 8 lines

Searching around the net, I find many other people having trouble with these cards and the ALSA-Linux2.6 combination. On one page, someone suggested that there were changes made between 2.4 and 2.6 to the ISA code that broke ALSA's detection routines.

I'm not sure what information might be needed in order to get this card working well once and for all, but if someone will let me know, I'd be happy to provide.

Does the OSS driver work in 2.6?

Quoted 3 lines

Thanks.

-Andy

cu
Adrian

Signature · 12 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed



-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
#2 lore raw
Quoted 1 line · Adrian Bunk

Does the OSS driver work in 2.6?

I was able to get the OSS module (opl3sa2) installed in 2.6 and I was able to get some nice hiss when doing 'cat /dev/urandom > /dev/dsp'. I used the following modprobe line:

modprobe opl3sa2 io=0x370 irq=11 dma=0 dma2=1 mss_io=0x530 isapnp=0

the isapnp=0 part appeared to be required. The following showed up in dmesg upon inserting the OSS module:

opl3sa2: Chipset version = 0x7
opl3sa2: Found OPL3-SA3 (YMF715E or YMF719E)

However, when I try to rmmod the opl3sa2 module, I get an Oops:

Unable to handle kernel NULL pointer dereference at virtual address 00000000
 printing eip:
c037ae01
*pde = 00000000
Oops: 0002 [#1]
PREEMPT
Modules linked in: opl3sa2 ad1848 mpu401 sound soundcore vfat fat
sd_mod usb_storage lp parport ide_scsi rtc
CPU:    0
EIP:    0060:[<c037ae01>]    Not tainted VLI
EFLAGS: 00010046   (2.6.12.2)

EIP is at wait_for_completion+0x71/0xf0
eax: cca91180 ebx: c52ea000 ecx: c52ebf20 edx: 00000000
esi: c52ea000 edi: cca9117c ebp: c52ebf40 esp: c52ebef4
ds: 007b es: 007b ss: 0068
Process rmmod (pid: 11277, threadinfo=c52ea000 task=c79ce550)

Stack: 00000000 c79ce550 c0113cf0 00000000 00000000 c52ea000 c0e899e0 c0209ffe
       00000001 c79ce550 c0113cf0 cca91180 00000114 00000001 00000000 cca91174
       00000114 00000001 00000000 c52ea000 cca8f964 cca91174 00000000 cca91200
Call Trace:
 [<c0113cf0>] default_wake_function+0x0/0x20
 [<c0209ffe>] kobject_put+0x1e/0x30
 [<c0113cf0>] default_wake_function+0x0/0x20
 [<cca8f964>] cleanup_opl3sa2+0x74/0x8c [opl3sa2]
 [<c0131bc8>] sys_delete_module+0x178/0x1b0
 [<c015b2c0>] sys_munmap+0x50/0x80
 [<c0103205>] syscall_call+0x7/0xb
Code: 00 00 8b 03 c7 45 d4 01 00 00 00 c7 45 bc f0 3c 11 c0 c7 45 dc

f0 3c 11 c0 89 45 b8 89 45 d8 8d 47 04 8b 50 04 89 45 e0 89 48 04 <89> 0a 89 55 e4 8d 76 00 8d bc 27 00 00 00 00 8b 03 c7 00 02 00
<6>note: rmmod[11277] exited with preempt_count 1

and now when I try to rmmod the module again, I get:

ERROR: Removing 'opl3sa2': Device or resource busy

So, yes, given the correct parameters, the OSS module (just like the ALSA module) works with the occasional Oops. I guess the Oopses are acceptable for now; it's the ALSA detection routines that are broken themselves or that are broken by something in the 2.6 kernel (remember that they work fine in 2.4).

Thanks.

-Andy

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_idt77&alloc_id492&op=click

#3 lore raw
In reply to Andrew Haninger
Quoted 10 lines · Andrew Haninger

Hello.

I have a 5 year old Gateway Solo 2500 that is currently running Linux 2.6.12.2. If I install ALSA and try to have alsaconf bruteforce-detect the OPL3SA2 sound card, it will say that it has detected it, but loading the modules will fail. If I install Linux 2.4 and recompile/rerun alsaconf, the detection works fine and the card works. Copying the configuration detected under 2.4 into a modprobe.conf on 2.6 allows me to use the card in 2.6 with occasional crashes (which might be due to suspend2).

Please, report this problem to our bug-tracking-system:

https://bugtrack.alsa-project.org/alsa-bug

We have already two similar reports #440 and #879. Please, provide us all info (working manual conf etc.)..

Jaroslav

-----
Jaroslav Kysela <[email protected]>
Linux Kernel Sound Maintainer
ALSA Project, SUSE Labs

#4 lore raw
Quoted 2 lines · Jaroslav Kysela

We have already two similar reports #440 and #879. Please, provide us all info (working manual conf etc.)..

Yes. #879 is the one that mentions that this might be a kernel bug (so I reported it here):

"I suspect this is kernel bug and it conserns ISA support. I would be happy if you kick some kernel developers buts because they have ignored my bugreport for quite long time now."

I don't know how correct/incorrect the above may be, but I figured it was possible.

I'll report what I know in bug #879.

Thanks.

-Andy

#5 lore raw
Quoted 1 line · Andrew Haninger

I'll report what I know in bug #879.

*Durr* I didn't scroll down. I've already reported my current semi-working (works+oopses) setup in that bug report

I'm reporting this here now because of the part of that bug report that suggests that it might actually be a kernel bug.

-Andy