NetBSD-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: [10.0_STABLE] Hard lock



Michael van Elst a écrit :
> joel.bertrand%systella.fr@localhost (=?UTF-8?Q?BERTRAND_Jo=c3=abl?=) writes:
> 
>> [ 24082.528401] cpu4[3392 slave]: hogging kernel lock
>> [ 24082.528401] Kernel lock error: _kernel_lock,266: spinout
> 
>> [ 24082.528401] ipi_msg_cpu_handler() at netbsd:ipi_msg_cpu_handler+0x68
>> [ 24082.528401] lock address : netbsd:kernel_lock
>> [ 24082.528401] type         : spin
>> [ 24082.528401] initialized  : netbsd:main+0x72
>> [ 24082.528401] shared holds :                  0 exclusive:
>>      1
>> [ 24082.528401] shares wanted:                  0 exclusive:
>>      5
>> [ 24082.528401] relevant cpu :                  2 last held:
>>      4
>> [ 24082.528401] relevant lwp : 0xffffaceebb7ad6c0 last held:
>> 0xffffaceec299d0c0
>> [ 24082.528401] last locked* : netbsd:tcp_rcvd_wrapper+0x1a
>> [ 24082.528401] unlocked     : netbsd:ipintr+0x1e7
>> [ 24082.528401] curcpu holds :                  0 wanted by:
>> 0xffffaceebb7ad6c0
> 
>> [ 24082.528401] panic: LOCKDEBUG: Kernel lock error: _kernel_lock,266:
>> spinout
> 
> 
> That's a different kind of locking problem. Not sure what triggers
> this in your case, but it reminds me about the symptoms of a forwarding
> loop in the network stack. It's only visible with LOCKDEBUG as that
> adds a 10 second timeout to the kernel lock.

	Maybe. But this panic occurs around 06:00 AM. Hard locks often occur
during night (between 02:00 and 07:00 AM). And I'm not sure that hard
lock on my system is related to iscsi itself. Maybe iscsi support raises
another locking problem.

Attachment: signature.asc
Description: OpenPGP digital signature



Home | Main Index | Thread Index | Old Index