kern/45648: Powering off unmounted SATA disk panics

[Hauke Fath <hf%spg.tu-darmstadt.de@localhost>]

>Number:         45648
>Category:       kern
>Synopsis:       Powering off unmounted SATA disk panics
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Nov 24 12:30:01 +0000 2011
>Originator:     Hauke Fath
>Release:        NetBSD 5.1_STABLE
>Organization:
TU Darmstadt
>Environment:
        
        
System: NetBSD venediger 5.1_STABLE NetBSD 5.1_STABLE (VENEDIGER) #0: Wed Feb 
23 21:05:10 CET 2011 
hf@Hochstuhl:/var/obj/netbsd-builds/5/i386/sys/arch/i386/compile/VENEDIGER i386
Architecture: i386
Machine: i386
>Description:

        A sata disk, powered up and connected to the machine, but not
        mounted (might not have been mounted ever, but since the
        machine had been up 150+ d, it is hard to tell) was powered
        off. A few hours later, the machine panicked with


piixide0:1:1: recal drive fault
wd0d: device fault reading fsbn 0 (wd0 bn 0; cn 0 tn 0 sn 0), retrying
NMI ... going to debugger
Stopped in pid 0.33 (system) at netbsd:lapic_delay+0x42: movl 
netbsd:__link_set_bufq_strats_sym_bufq_strat_dummy+0x1684,%edi
db{1}> 


        The above is from serial console - no stack trace, since the
        machine had to be brought up asap.

>How-To-Repeat:

        Power off an unused (as in: not mounted)  sata disk.

        -- I don't know if it is that easy... 
 
>Fix:
        Something in the kernel appears to have hung on to the disk,
        and something else appears to have considered the resulting
        error 'fatal'. Both should not have happened.

>Unformatted: