As an experiment, I tried a raidctl -f and then a raidctl -R on one
of my RAID components. The raidctl -R paniced... it said "quiescence
reached" and then right after that "panic: out of space in kmem_map".

This happens both with netbsd 1.5 and with -current as of today.

The array is working fine in degraded mode, but obviously I'd like
to get it reconstructed if I can :-) Any ideas?

In case it helps, here's my raid0.conf:

---cut here---
START array
1 3 0

START disks
/dev/sd0g
/dev/sd1g
/dev/sd2g

START layout
16 1 1 5

START queue
fifo 100
---cut here---

Each component is exactly 16 gigabytes, or 33554432 sectors. The
machine has 224 megs of physical RAM, and the raidctl -R was done in
single-user mode. The parity was clean and all components were optimal
as of when I did the raidctl -f.

I can put crash dumps up for ftp if anyone's curious.

Thanks!

	~Ben