Re: -current on OPENRD doesn't boot

[SAITOH Masanobu <msaitoh%execsw.org@localhost>]

(2012/09/18 9:34), SAITOH Masanobu wrote:
> (2012/09/16 20:27), SAITOH Masanobu wrote:
>>  Hi.
>>
>>  My OPEN-RD with latest -current have not booted since 2 days ago.
> 
> With SATA disk, it doesn't boot. Without SATA disk (i.e. NFS diskless),
> it boots.
> 
> Stack trace is:
> 
>> root device (default mvgbe0): wd0
>> dump device (default wd0b):
>> file system (default ffs): ffs
>> root on wd0a dumps on wd0b
>> Mutex error: mutex_vector_enter: locking against myself
>>
>> lock address : 0x00000000c0c8efd8
>> current cpu  :                  0
>> current lwp  : 0x00000000c2a4fd20
>> owner field  : 0x0000000001000700 wait/spin:                0/1
>>
>> panic: lock error
>> Stopped in pid 0.2 (system) at  netbsd:cpu_Debugger+0x4:        bx      r14
>> db> trace
>> netbsd:vpanic+0x10
>>         scp=0xc09f6f84 rlv=0xc09f721c (netbsd:printf_nolog)
>>         rsp=0xc0cb8bc0 rfp=0xc0cb8bd4
>>         r7=0xc0b4d3bc
>> netbsd:panic+0x14
>>         scp=0xc09f720c rlv=0xc09ef9d0 (netbsd:lockdebug_abort+0x78)
>>         rsp=0xc0cb8be8 rfp=0xc0cb8c10
>> netbsd:lockdebug_abort+0x10
>>         scp=0xc09ef968 rlv=0xc09157d4 (netbsd:mutex_abort+0x2c)
>>         rsp=0xc0cb8c14 rfp=0xc0cb8c20
>>         r7=0x00000080 r6=0xc0c7df8c
>>         r5=0xc0c8efd8 r4=0xc0c8efd0
>> netbsd:mutex_abort+0x10
>>         scp=0xc09157b8 rlv=0xc09159f0 (netbsd:mutex_enter+0x68)
>>         rsp=0xc0cb8c24 rfp=0xc0cb8c50
>> netbsd:mutex_enter+0x10
>>         scp=0xc0915998 rlv=0xc093144c (netbsd:tc_ticktock+0x3c)
>>         rsp=0xc0cb8c54 rfp=0xc0cb8c64
>>         r10=0x00000000 r9=0x00000006
>>         r8=0x00000000 r7=0x00000080 r6=0xc0c7df8c r5=0xc2a4fd20
>>         r4=0xc0c8efd0
>> netbsd:tc_ticktock+0x10
>>         scp=0xc0931420 rlv=0xc08fcfe0 (netbsd:hardclock+0xc0)
>>         rsp=0xc0cb8c68 rfp=0xc0cb8c7c
>>         r4=0xc0cb8d40
>> netbsd:hardclock+0x10
>>         scp=0xc08fcf30 rlv=0xc0954d58 (netbsd:clockhandler+0x10)
>>         rsp=0xc0cb8c80 rfp=0xc0cb8c8c
>>         r5=0xc0c7df84 r4=0xc29d0458
>> netbsd:clockhandler+0x10
>>         scp=0xc0954d58 rlv=0xc09a581c (netbsd:pic_dispatch+0x34)
>>         rsp=0xc0cb8c90 rfp=0xc0cb8cac
>> netbsd:pic_dispatch+0x10
>>         scp=0xc09a57f8 rlv=0xc09a5ba8 (netbsd:pic_do_pending_ints+0x2d4)
>>         rsp=0xc0cb8cb0 rfp=0xc0cb8d18
>>         r6=0xc0c7df8c r5=0xc0c7df84
>>         r4=0x00000001
>> netbsd:pic_do_pending_ints+0x10
>>         scp=0xc09a58e4 rlv=0xc095437c (netbsd:mvsoc_irq_handler+0x74)
>>         rsp=0xc0cb8d1c rfp=0xc0cb8d3c
>>         r10=0x00000000 r9=0x00000001
>>         r8=0x00000000 r7=0x00000132 r6=0xc0cb8d40 r5=0x00000000
>>         r4=0x00000001
>> netbsd:mvsoc_irq_handler+0x10
>>         scp=0xc0954318 rlv=0xc08f9eec (netbsd:irq_entry+0x6c)
>>         rsp=0xc0cb8d40 rfp=0xc0cb8dbc
>>         r8=0x00000008 r7=0x00000013
>>         r6=0x00000001 r5=0xc0c872e4 r4=0xc0b9314c
>> 10
>>         scp=0x00000010 rlv=0x00000006 (6)
>>         rsp=0x00000001 rfp=0xc2d1df88
>> 8
>>         scp=0x00000008 rlv=0x0000ffff (ffff)
>>         rsp=0x0000ffff rfp=0x0000ffff
>> Non-emulated page fault with intr_depth > 0
>> Fatal kernel mode data abort: 'Translation Fault (S)'
>> trapframe: 0xc0cb8498
>> FSR=00000005, FAR=fffffffc, spsr=60000053
>> r0 =0000001e, r1 =00000000, r2 =fffffff4, r3 =00000000
>> r4 =00000008, r5 =c2d1df88, r6 =c09f78f8, r7 =e59ff114
>> r8 =c0cb8d18, r9 =0000fff3, r10=00000001, r11=c0cb8968
>> r12=c0cb8440, ssp=c0cb84ec, slr=c09f6c78, pc =c0842628
>>
>> Faulted in DDB; continuing...
>> db>


 skrll adviced me to add LOCKDEBUG option. The result is:

> Mutex error: mutex_vector_enter: locking against myself
> 
> lock address : 0x00000000c0c8f118 type     :               spin
> initialized  : 0x00000000c09321e8
> shared holds :                  0 exclusive:                  0
> shares wanted:                  0 exclusive:                  1
> current cpu  :                  0 last held:                  0
> current lwp  : 0x00000000c0c7c7c0 last held: 000000000000000000
> last locked  : 0x00000000c093217c unlocked*: 0x00000000c09321b8
> owner field  : 0x0000000001000700 wait/spin:                0/1
> 
> panic: LOCKDEBUG
> Stopped in pid 0.1 (system) at  netbsd:cpu_Debugger+0x4:        bx      r14
> db>

Address c09321xx is in kern_tc.c::tc_ticktock().
mvsoc_intr.c have not changed for 1 month and the definition of
the mask register and cause register have not changed for a long
time.

 Are lock/unlock stuff or atomic operations broken?

-- 
-----------------------------------------------
                SAITOH Masanobu (msaitoh%execsw.org@localhost
                                 msaitoh%netbsd.org@localhost)