Over the weekend, I was using my Sun-3/110 to attack the ffs_alloccgblk
panic, and when I was working on a vnd filesystem I got

panic: locking against myself

I reported this on port-sun3, because (a) vnd is commented out in the
sun3 GENERIC config and (b) I've never had any trouble using vnd
"disks" on the sparc port, thus suspecting there was something
sun3-specific about it.  However, gwr said

> That looks like unwanted recursion to me, and is probably due to
> something in the common code.  Please repost on current-users

so here it is.

This is a ten-finger copy of the traceback at the time of the crash.

panic: locking against myself
Stopped at      _Debugger+0x6:  unlk    a6
db> trace
_Debugger(e01aa84,e0875aa,eda7934,10,eda7940) + 6
_panic(e0875aa,e5758f0,e75e7d4,e8d4000,eda79a8) + 34
_ufs_lock(eda79a0) + 64
_vndstrategy(e75e7d4,eda79d4,e0265ca,eda79cc,2004) + f6
_spec_strategy(eda79cc) + 24
_bwrite(e75e7d4,eda79f4,e0266e2,eda79ec,e0bdba0) + b8
_vn_bwrite(eda79ec,e0bdba0,e75e7d4,eda7a08,e026c36) + c
_bawrite(e75e7d4) + 36
_getnewbuf(0,0) + 7a
_getblk(e5ab400,ffffcff4,2000,0,0) + c4
_ufs_bmaparray(e5ab400,31f0,eda7b28,0,0) + 15a
_ufs_bmap(eda7b0c) + 3c
_vndstrategy(e75c104,eda7b64,e0265ca,eda7b5c,2004) + 13e
_spec_strategy(eda7b5c) + 24
_bwrite(e75c104,eda7b84,e0266e2,eda7b7c,e0bdba0) + b8
_vn_bwrite(eda7b7c,e0bdba0,e75c104,eda7b98,e026c36) + c
_bawrite(e75c104) + 36
_getnewbuf(0,0) + 7a
_getblk(e5a4400,0,1400,0,0) + c4
_bread(e5a4400,0,1400,ffffffff,eda7c28) + 20
_ffs_blkatoff(eda7ca0) + 82
_ufs_lookup(eda7d04) + 3f0
_lookup(eda7ea0) + 234
_namei(eda7ea0) + 16a
_check_exec(e58e800,eda7ee8,40,3f,0) + 24
_sys_execve(e58e800,eda7f84,eda7f7c) + c2
_syscall(3b) + 13a
_trap0() + e
db> c
syncing disks... panic: locking against myself
Stopped at      _Debugger+0x6:  unlk    a6
db> c
dumping not supported yet :)
Kernel rebooting...
Boot: sd(0,0,0)
Invalid Page Bus Error:
  Vaddr: 0E58E7FF, Paddr: 000007FF, Type 0, Read, FC 5, Size 1 at 0x00004014
>

I'm hoping someone has some idea what's wrong.  I don't really.

The system this happend on was entirely built from sources supped in
the AM on Feb 27, around 08:30-09:00ish Eastern.  It seems more or less
repeatable; it's happened twice for two now.  If anyone has any idea
what I might usefully look at when it happens, speak up and I can
probably make it happen again and investigate.

					der Mouse

			    mouse@collatz.mcrcim.mcgill.edu