FUSE crashes on i386 and amd64

[Tom Ivar Helbekkmo <tih%hamartun.priv.no@localhost>]

I'm experimenting with MooseFS on NetBSD/i386 and /amd64, current as of
September 9th.  The server side seems to work fine, with everything
behaving just as it should when using a Linux client.  The NetBSD
client, however, crashes the kernel.  Here it is on amd64:

barsoom# crash -M netbsd.25.core -N netbsd.25
Crash version 7.99.1, image version 7.99.1.
System panicked: trap
Backtrace from time of crash is available.
crash> bt
_KERNEL_OPT_NARCNET() at 0
_KERNEL_OPT_NARCNET() at 0
vpanic() at vpanic+0x145
snprintf() at snprintf
startlwp() at startlwp
calltrap() at calltrap+0x19
puffs_cookie2vnode() at puffs_cookie2vnode+0x31
puffs_sop_thread() at puffs_sop_thread+0xeb

...and this is i386:

athene# crash -M netbsd.15.core -N netbsd.15
Crash version 7.99.1, image version 7.99.1.
System panicked: kernel diagnostic assertion "vp->v_vflag & VV_ROOT"
failed: file "/usr/src/sys/fs/puffs/puffs_vnops.c", line 1431 
Backtrace from time of crash is available.
crash> bt
_KERNEL_OPT_NARCNET(0,104,c063d1c5,8,c0c7039d,0,104,c0c53d44,dca94e5c,dca94e40)
at 0
_end(104,0,c0c53d44,dca94e5c,c4d57000,c3fb442c,1,dca94e50,c0a05e7b,c0c53d44)
at dca94e5c
vpanic(c0c53d44,dca94e5c,dca94e78,c073a385,c0c53d44,c0c53f10,c0cd761d,c0cd74b4,597,c3fb442c)
at vpanic+0x131
kern_assert(c0c53d44,c0c53f10,c0cd761d,c0cd74b4,597,c3fb442c,c4d54a80,c3fb442c,dca94e98,c099eb60)
at kern_assert+0x23
puffs_vnop_reclaim(dca94e8c,0,0,c0c432d0,c3fb442c,1,dca94ed0,c097ac58,c3fb442c,1)
at puffs_vnop_reclaim+0x130
VOP_RECLAIM(c3fb442c,1,ffffffff,c4d54a80,0,0,0,0,c0c432f4,c3fb442c) at
VOP_RECLAIM+0x4a
vclean(c5784a40,dca94ef3,c5784a40,c3f9dd80,c3ef0008,0,1000004,c4d57000,c4d57000,dca94f7c)
at vclean+0xd8
vrelel(c3fb442c,bb90e180,4,dca94f7c,c0cd73f7,c0ec15d0,c4d54a80,c339dd40,c4d57fd4,c4d57000)
at vrelel+0x5ae
puffs_cookie2vnode(c4d57000,bb90e180,dca94f7c,0,c4d54a80,c4cd5020,c4d54a80,dca94f88,c05b4f8e,c4d54a80)
at puffs_cookie2vnode+0x113
puffs_sop_thread(c4d57000,10c0000,10ca000,0,c0100304,0,0,0,0,0) at
puffs_sop_thread+0xf5

I've looked at the PUFFS code, but don't understand it sufficiently well
to see what's wrong.  Anyone with an idea for something I can try?

-tih
-- 
It doesn't matter how beautiful your theory is, it doesn't matter how smart
you are. If it doesn't agree with experiment, it's wrong.  -Richard Feynman