kern/58776: RAIDframe panic on I/O error during reconstruction

[manu%netbsd.org@localhost]

>Number:         58776
>Category:       kern
>Synopsis:       RAIDframe panic on I/O error during reconstruction
>Confidential:   no
>Severity:       serious
>Priority:       low
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Oct 25 08:10:00 +0000 2024
>Originator:     Emmanuel Dreyfus
>Release:        NetBSD 10.0
>Organization:
NetBSD
>Environment:
NetBSD-10.0/i386
>Description:
On NetBSD-10, a RAID 1 with two disks get un unrecoverable I/O error
during reconstruction, which sometime happens. But then it panics,
which is worse that just failing a RAID reconstruction.

panic: trap     
cpu0: Begin traceback...
vpanic(c0d6708c,df0fcdd0,df0fce8c,c012fd98,c0d6708c,df0fce98,df0fce98,ea,df0fb2c0,210206) at netbsd:vpanic+0x196        
panic(c0d6708c,df0fce98,df0fce98,ea,df0fb2c0,210206,fffffff0,0,0,c42716a0) at netbsd:panic+0x18
trap() at netbsd:trap+0xd51
--- trap (number 6) ---
                                                              
mutex_vector_enter(c6f02c7c,1,1,c6dd6f44,c6dd6f44,c670316c,c6dd7308,df0fcf78,c078dc03,c670316c) at netbsd:mutex_vector_enter+0x91
rf_CauseReconEvent(c670316c,1,c6dd6f44,9,c6703244,df0fcf9c,c077e310,c6dd6f44,5,5) at netbsd:rf_CauseReconEvent+0x58
ReconWriteDoneProc(c6dd6f44,5,5,c6703248,c670316c,c077e271,c6c19c00,0,c0102011,c670316c) at netbsd:ReconWriteDoneProc+0x79
rf_RaidIOThread(c670316c,43e6000,43fb000,0,c01005a8,0,0,0,0,0) at netbsd:rf_RaidIOThread+0x9f

>How-To-Repeat:
	raidctl -R /dev/dk3 raid1
	But it is probably very specific to the hardware state.
>Fix:
	None known yet