Re: panic: ffs_valloc: dup alloc

[gson%gson.org@localhost (Andreas Gustafsson)]

David Holland wrote:
> Do you have any ideas how one might set about replicating it?

This may not be exactly what you are looking for, but I'll just note
that I can reproduce the "ffs_valloc: dup alloc" panic at will by
mounting a filesystem left in a dirty state by the crash of kern/38762
and creating a new file on it.

I have put a compressed image of a dirty filesystem that will
trigger the panic at

  http://www.gson.org/netbsd/bugs/dupalloc/test.fs.bin.gz

This will uncompress into a 1 megabyte hard disk image.  Here's
a typescript from a qemu session where that image was mapped
to the secondary hard drive using "qemu -hdb test.fs.bin":

  # mount /dev/wd1a /mnt
  /dev/wd1a: file system not clean (fs_clean=0x4); please fsck(8)
  /dev/wd1a: lost blocks 0 files 0
  # >/mnt/foobar
  dmode 81a4 mode 81a4 dgen 320f8b3d gen 320f8b3d
  size 0 blocks 0
  ino 3 ipref 2
  panic: ffs_valloc: dup alloc
  fatal breakpoint trap in supervisor mode
  trap type 1 code 0 eip c0236ad4 cs 8 eflags 246 cr2 bbbe2080 ilevel 0
  Stopped in pid 296.1 (sh) at    netbsd:breakpoint+0x4:  popl    %ebp
  db{0}> 

-- 
Andreas Gustafsson, gson%gson.org@localhost