CD eject -> panic on ALPHA8

[Kenneth D. Merry]

On Sun, Oct 07, 2018 at 22:43:09 +0200, Harry Schmalzbauer wrote:
> Hello,
> 
> like recorded in https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=232044
> the following panic happens after ejecting ODD media on 12.0_ALPHA8 #0 
> r339222M:
> 
> panic: sleepq_add: td 0xfffff80002760580 to sleep on wchan 0xfffff8000a905048 with sleeping prohibited
> cpuid = 0
> time = 1538941032
> KDB: stack backtrace:
> db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfffffe00004e2610
> vpanic() at vpanic+0x1a3/frame 0xfffffe00004e2670
> panic() at panic+0x43/frame 0xfffffe00004e26d0
> sleepq_add() at sleepq_add+0x32c/frame 0xfffffe00004e2720
> _sleep() at _sleep+0x261/frame 0xfffffe00004e27c0
> cam_periph_runccb() at cam_periph_runccb+0x18d/frame 0xfffffe00004e2910
> cdcheckmedia() at cdcheckmedia+0x8a/frame 0xfffffe00004e2990
> cdstrategy() at cdstrategy+0x55/frame 0xfffffe00004e29c0
> g_disk_start() at g_disk_start+0x334/frame 0xfffffe00004e2a20
> g_io_schedule_down() at g_io_schedule_down+0x105/frame 0xfffffe00004e2a60
> g_down_procbody() at g_down_procbody+0x6d/frame 0xfffffe00004e2a70
> fork_exit() at fork_exit+0x84/frame 0xfffffe00004e2ab0
> fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe00004e2ab0
> --- trap 0, rip = 0, rsp = 0, rbp = 0 ---
> KDB: enter: panic
> 
> #10 0xffffffff80592083 in panic (fmt=<value optimized out>)
>      at /usr/src/sys/kern/kern_shutdown.c:799
> #11 0xffffffff805e748c in sleepq_add (wchan=0xfffff8000a905048, lock=0xfffff8000a94dcd0, wmesg=0xffffffff809cdf89 "cbwait",
>      flags=0, queue=0) at /usr/src/sys/kern/subr_sleepqueue.c:319
> #12 0xffffffff8059ccd1 in _sleep (ident=0xfffff8000a905048, lock=<value optimized out>, priority=92,
>      wmesg=0xffffffff809cdf89 "cbwait", sbt=<value optimized out>, pr=0, flags=256)
>      at /usr/src/sys/kern/kern_synch.c:201
> #13 0xffffffff802ba7ed in cam_periph_runccb (ccb=0xfffff8000a905000, error_routine=0xffffffff802e10e0 <cderror>,
>      camflags=CAM_RETRY_SELTO, sense_flags=3, ds=0xfffff8000aa7fc60)
>      at /usr/src/sys/cam/cam_periph.c:1141
> #14 0xffffffff802e09ea in cdcheckmedia (periph=0xfffff8000ad44300)
>      at /usr/src/sys/cam/scsi/scsi_cd.c:821
> #15 0xffffffff802dd125 in cdstrategy (bp=0xfffff8000aa2f600) at /usr/src/sys/cam/scsi/scsi_cd.c:861
> #16 0xffffffff8050eb54 in g_disk_start (bp=<value optimized out>)
>      at /usr/src/sys/geom/geom_disk.c:478
> #17 0xffffffff80512575 in g_io_schedule_down (tp=<value optimized out>)
>      at /usr/src/sys/geom/geom_io.c:893
> #18 0xffffffff80512d6d in g_down_procbody (arg=<value optimized out>)
>      at /usr/src/sys/geom/geom_kern.c:111
> #19 0xffffffff8055a5d4 in fork_exit (callout=0xffffffff80512d00 <g_down_procbody>, arg=0x0, frame=0xfffffe00004e2ac0)
>      at /usr/src/sys/kern/kern_fork.c:1057
> #20 0xffffffff808b917e in fork_trampoline () at /usr/src/sys/amd64/amd64/exception.S:993
> #21 0x0000000000000000 in ?? ()

Actually, there is already a bug outstanding for this issue (219857), and
I've posted a patch, but evidently it has issues.  (I closed yours as a
duplicate, so we can keep the history in one place.)

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=219857

I need someone to help debug the patch that's attached to that bug.  Care
to give it a try?

Thanks,

Ken
-- 
Kenneth Merry
ken at FreeBSD.ORG