Hello!
The spca50x driver (Project located at sourceforge.net) has a
particularly anoying bug.
In essence, when the spca50x driver releases memory, the system hangs
(Freezes). The
system does not crash evertime, but about 1 in 3 times. What I am
interested in doing is
fixing the bug (It's never been assigned to anyone since it was
discovered).
Anyway, here is a copy of the kernel panic:
kernel BUG at panic.o:161
invalid operand 0000 2.4.19-4GB at <actual Date>
CPU 0
EIP 0010:[<c011e89f>] not tainted
EIFLAGS 00010286
eax: 00000025 ebx:00461000 ecx:c0317f04 edx:c029c740
esi: d9561000 edi:00000000 ebp:d9561000 esp:c0317f00
dc. 0018 es:0018 ss:0018
Stack: c029c740 00000041 c4ba19c4 00000041 c4bc5a04
c4bc5000 c03616e0 00000001 c4ba499f d9561000
004b0020 c4bc5714 c4ba49e4 c012731c c4bc5000
c0317f4c c0317f4c 00000000 c0341f40 fffffffe
call trace
c4ba19c4 c4ba499f c4ba4a4e c4ba49e4 c012731c c0123612
c012352b c0123379 c010a251 c0106f70 c010c758 c0106f70
c0106f94 c0106fe7 c0105000
modules spca50x:<c4ba0060>:<c4bb5300>
code 0f 06 00 46 bf 29 c0 58 5a 8d b4 26 00 00 00 00 eb fe 90
<0> kernel panic Aiee, killing interrupt handler
in interrupt handler - syncing
The system log says:
Apr 30 14:16:44 115 kernel: drivers/usb/spca50x.c:
[spca50x_dealloc:2988] deallocating buffer memory later
Apr 30 14:16:44 115 kernel: drivers/usb/spca50x.c:
[spca50x_dealloc:3006] leaving
Apr 30 14:20:54 115 syslogd 1.4.1: restart.
It's been about 10 years since I wrote any drivers. Further, the drivers
I have written were for another flavor of UNIX and not Linux. What I am
wondering is if any driver developers who on the list have any thoughts
on what is causing memory not to be released.
Any and all responses are welcome.
Best regards
Marvin Dickens
Alpharetta, Georgia USA
