Hi,

I had been using Chuck Silvers' pmap patch, and haven't had a problem in 
weeks of stress testing. Yesterday morning, the system's power got yanked 
while in the middle of running stuff, and since then, I've gotten two 
panics like this:


pool_get: pmap_upvopl: items on itemlist, nitems 0
panic: pool_get: nitems inconsistent
Begin traceback...
0xd72b5970: at pool_get+0x17c
0xd72b59a0: at pmap_pvo_enter+0x108
0xd72b59f0: at pmap_kenter_pa+0xac
0xd72b5a10: at uvm_pagermapin+0x1ac
0xd72b5a40: at genfs_gop_write+0x114
0xd72b5b00: at genfs_putpages+0x6e8
0xd72b5c50: at ffs_putpages+0x3c
0xd72b5c70: at ffs_write+0x548
0xd72b5db0: at vn_write+0x150
0xd72b5e40: at dofilewrite+0xc4
0xd72b5eb0: at sys_write+0x8c
0xd72b5ed0: at syscall_plain+0xe0
0xd72b5f40: user SC trap #4 by 0x41c03428: srr1=0xd032
             r1=0xffffdee0 cr=0x88004082 xer=0 ctr=0
End traceback...
syncing disks...


And...


pool_get: pmap_mpvopl: curpage NULL, nitems 2
panic: pool_get: nitems inconsistent
Begin traceback...
0xd835b7d0: at pool_get+0x17c
0xd835b800: at pmap_pvo_enter+0x108
0xd835b850: at pmap_enter+0x198
0xd835b880: at ubc_fault+0x37c
0xd835b930: at uvm_fault+0x1068
0xd835ba60: at trap+0x2c0
0xd835bae0: kernel DSI write trap @ 0xd585a000 by memcpy+0x90: srr1=0x9032
             r1=0xd835bba0 cr=0x48004024 xer=0x20000000 ctr=0x800 
dsisr=0x4200000
0xd835bba0: at _SDA_BASE_+0xf64
0xd835bbb0: at pmap_pvo_remove+0x240
0xd835bc40: at uiomove+0x1c4
0xd835bc70: at ffs_write+0x450
0xd835bdb0: at vn_write+0x150
0xd835be40: at dofilewrite+0xc4
0xd835beb0: at sys_write+0x8c
0xd835bed0: at syscall_plain+0xe0
0xd835bf40: user SC trap #4 by 0x41dc3438: srr1=0xd032
             r1=0xffffcbe0 cr=0x88004022 xer=0 ctr=0
End traceback...
syncing disks... Stopped in pid 13.1 (aiodoned) at 
netbsd:cpu_Debugger+0xr
0, r1, 0x14
db> reboot 0x4


Two things - anyone know what might be causing this? Also, am I not 
getting a crashdump because I have more memory than swap?

Thanks,
John Klos