This panic is on a PowerMac 9600 with a 750GX accelerator, 1.5 gigs of 
memory, and 2 gigs of swap. It's been completely stable for months and 
months, but recently started getting extremely heavy use which includes 
250 running Apache processes with mod_php plus lots of MySQL work, so it's 
been going heavily into swap. I have no idea how far into swap it was when 
it paniced, but it was nearly 800 megs into swap in the morning.


trap: pid 6.1 (pagedaemon): kernel MCHK trap @ 0x2674b4 (SRR1=0x49030)
panic: trap
Begin traceback...
0xd5a34d40: at trap+0xec
0xd5a34dc0: kernel MCHK trap by pool_reclaim+0xb0: srr1=0x49030
             r1=0xd5a34e80 cr=0x44000028 xer=0 ctr=0x268564
0xd5a34e80: at pool_reclaim+0x7c
0xd5a34ee0: at pool_drain+0x60
0xd5a34f00: at uvm_pageout+0x164
0xd5a34f40: at cpu_switchto+0x44
0xd5a34f50: at ADBDevTable+0x1085bab4
saved LR(0x3) is invalid.End traceback...
syncing disks... done
dumpsys: TBD
rebooting


Any ideas on what's allowing this to happen?

It's running netbsd-2, BTW.

Thanks,
John Klos