Description of problem: Every time I shut my P4 Prescott system down, I get a kernel oops that appears to be caused by ALSA. Version-Release number of selected component (if applicable): 2.6.16-1.2133_FC5smp How reproducible: Every time. Steps to Reproduce: 1. shutdown machine 2. observe oops. Actual results: Below is the bits of the oops I managed to copy down. Note: I don't have a serial cable, so what you see below has been hand-typed. The text delimited by "###" is my comments - I didn't get all the info down I'm afraid. -------------------------------------------------------------------------- CPU: 0 EIP: 0060:[<f8925520>] Not tainted VLI EFLAGS: 00010096 (2.6.16-1.2133_FC5smp #1) EIP is at snd_seq_delete_all_ports+0x6c/0x170 [snd_seq] eax: 00000044 ebx: f722a248 ecx: f3954f18 edx: f892683c esi: f722a248 edi: f7df83c0 ebp: f7df843c esp: f3954f14 Process rmmod #### pid details and context ### Stack: ### LIST OF ADDRESSES ONLY ### Call Trace: [<f88f0206>] seq_free_client1+0x8/0x90 [snd_seq] [<f88f02da>] seq_free_client+0x4c/0x82 [snd_seq] [<f88f2732>] snd_seq_delete_kernel_client+0x1a/0x2c [snd_seq] [<c013c3c2>] sys_delete_module=0x1cc/0x20c do_syscall_trace syscall_call #### something times out here, and the screen changes to a new dump: ### <3> Debug: sleeping function called from invalid context at include/linux/rwsem.h:43 ### a different stack trace here #### sysscall_cal+0x7/0xb /etc/rc0.d/K14alsasound: line 215: 2554 Segmentation fault /sbin/rmmod `echo $line | cut -d ' ' -f 1` -------------------------------------------------------------------------- Expected results: Clean shutdown. Additional info:
Note: 2.6.16-1.2133_FC5 (non-smp) also suffers from the same problem. However, the problem is not seen using kernel 2.6.16-1.2122_FC5smp.
My soundcard is reported as the following using 'lspci': 00:14.5 Multimedia audio controller: ATI Technologies Inc IXP150 AC'97 Audio Controller.
The other point to note is that very occasionally (using the 2133 kernel), my system will start pumping white noise out of the speakers and lock up hard. A kernel buffer overflow issue related to the alsa rmmod maybe?
I am still having the same problem with kernel 2.6.17-1.2139_FC5smp.
Can anybody comment on this please?
Well, somebody obviously tweaked something - the problem is no longer present in kernel 2.6.17-1.2145_FC5smp.
You can close this bug - it's no longer a problem!