> On my machine it works fine..
> Did you load other sound stuff (like sb.o and adlib_card.o) before
> loading awe_wave.o?
I have sb and adlib compiled into the kernel, but I use awe_wave as a
module, just out of habit (in earlier 2.1.* kernels it only worked that
way).
Any ideas about the "bang" that it makes when the module loads?
I can replicate the problem in this manner -
1. run xmp using normal /dev/sound with associated drivers.
2. load the awe_wave module
3. run xmp again, using AWE support
4. remove awe_wave module from the kernel using rmmod
5. run xmp. This time, it'll give the error.
6. run xmp again. It won't give an error. Quite strange.
This is the error I get on dmesg when I do this:
Unable to handle kernel NULL pointer dereference at virtual address 00000094
current->tss.cr3 = 00d95000, %cr3 = 00d95000
*pde = 00000000
Oops: 0000
CPU: 1
EIP: 0010:[<c01ce9b5>]
EFLAGS: 00010246
eax: 00000000 ebx: 00000001 ecx: 00000001 edx: 00000000
esi: 00000003 edi: 00000001 ebp: 00000001 esp: c39f9ef0
ds: 0018 es: 0018 ss: 0018
Process xmp (pid: 566, process nr: 47, stackpage=c39f9000)
Stack: c01cd360 c01cd294 00000000 00000001 00000001 00000000 c01d0647 00000001
c5f18560 00000000 c5f18560 c29f6ca0 c358c00e 00000001 c012dfa8 c0fe50e0
c01c6b93 c56cec34 c5f18560 00000000 c5f18560 c56cec34 c29f6ca0 c012dc46
Call Trace: [<c01cd360>] [<c01cd294>] [<c01d0647>] [<c012dfa8>] [<c01c6b93>]
[<c012dc46>]
[<c012756e>]
[<c01e3425>] [<c0127892>] [<c0126612>] [<c012677f>] [<c0108c00>]
Code: 8b 80 94 00 00 00 89 4c 24 1c ff d0 83 c4 10 8b 4c 24 0c 85
As for the other things, I am using a completely stock 2.2.1 kernel, with
no changes whatsoever. I'm compiling 2.2.2 as I write this.
Zack
#-----------------------------------------------------------------#
# Zack Williams [EMAIL PROTECTED] http://www.u.arizona.edu/~zdw #
# #
# Linux is like an arcwelder - it is insanely powerful in the #
# right hands, but most people get by with JB Weld #
#-----------------------------------------------------------------#
