[Bug 209099] ata2: already running! panic: bad link elm 0xfffff80003b7e6a0 prev->next != elm

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=209099

daryl at ci.com.au changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |daryl at ci.com.au

--- Comment #5 from daryl at ci.com.au ---
I have an Asus P8B-X motherboard with 16G memory and an Intel 530 Series
120G SSD as the boot drive. We are running FreeBSD 10.3-STABLE #3 r300822.

I believe I have this same issue. I was using salt minion to update this
machine and it hung (several times). Prior to the hang I would always see the
console message:

ata2: already running!

The machine would continue to run for a short period of time and then become
unresponsive. A few minutes later I might get a panic otherwise it would just 
be in a hung state.

I was able to pull apart the minion and found it was camcontrol causing the
issue. I create a script that looped over the command:

camcontrol identify ada0 

and I am able to cause the system to either hang or panic regularly.
Using the kernel debugger I was able to glean this information:

spin lock 0xffffffff816f17e0 (sleepq chain) held by 0xfffff800g
panic: spin lock held too long                                                  
cpuid = 6                                                                       
KDB: stack backtrace:                                                           
db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfffffe046562e800  
kdb_backtrace() at kdb_backtrace+0x39/frame 0xfffffe046562e8b0                  
vpanic() at vpanic+0x126/frame 0xfffffe046562e8f0                               
panic() at panic+0x43/frame 0xfffffe046562e950                                  
_mtx_lock_spin_cookie() at _mtx_lock_spin_cookie+0x287/frame 0xfffffe046562e9c0 
wakeup() at wakeup+0xf/frame 0xfffffe046562e9e0                                 
vnlru_proc() at vnlru_proc+0x157/frame 0xfffffe046562ea70                       
fork_exit() at fork_exit+0x9a/frame 0xfffffe046562eab0                          
fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe046562eab0               
--- trap 0, rip = 0, rsp = 0, rbp = 0 ---                                       
KDB: enter: panic                                                               
[ thread pid 20 tid 100078 ]                                                    
Stopped at      kdb_enter+0x3e: movq    $0,kdb_why

If salt is disabled and I dont run camcontrol commands, this machine is stable.
We have several other machines with different mother boards using the same
kernel that dont have this problem.

-- 
You are receiving this mail because:
You are on the CC list for the bug.