Hmm.. that last patch didn't do it. I've noticed some errors on the
console before the lockup:
unexpected md driver lock: 0xe1813900: type VREG, usecount 2, writecount 1, refcount
3871, flags (VOBJBUF)
tag VT_UFS, ino 4, on dev da0s1h (13, 131079) lock type inode: EXCL (count 1
Ok, I think I've whacked this one. Try this patch and see if it
fixes your buf_daemon() lockups. The patch also fixes the
double-data-caching that occurs with file-backed MD.
If this works for you, Mark, I'll commit it and probably also MFC it.
I'll also be able to apply th
Hmm. Well, bufdaemon is doing the right thing by blocking
in that loop. The problem appears to be related to the
double-I/O that is occuring due to the way MD filesystems
work. A vnode-backed MD takes a strategy call and turns
around and issues a high-level filesystem write.
