ZFS related kernel panic

[a.smith at ukgrid.net]

Hi,

   I am experiencing kernel panics when running a zfs receive. And  
unfortunately Im also unable to get a core dump written as its  
freezing at the point where its meant to write the dump to disk.
Initially I was on a FreeBSD 8.1-RELEASE i386 server with 2GB RAM  
running with an SIIS eSATA card and two eSATA disks attached (using a  
single cable using port multiplier).
The server with the primary copy of the data (that does the zfs sends)  
is running on amd64 OS with 4GB RAM, so I decided as this was ZFS  
related the quickest probable fix would be to go to amd64 with 4GB RAM  
on the receiving server too. I have now done that, Ive moved the eSATA  
cards and disks to a brand new server, and I am still seeing kernel  
panics, and Im still unable to write the dump to disk.
I have a Dell graphical remote console (no serial console), and the  
info on screen when the system crashes is like:

siisch0: Timeout on slot 23
siisch0: siis_timeout is 00040000 ss 7fffffff rs 7fffffff es 00000000  
sts 801d20
siisch0  ..waiting for slots 7f7fffff

Fatal trap 12: page fault while in kernel mode
cpuid = 0; apic id = 00
fault code = supervisor read, page not present
instruction pointer = 0x20:0xffffffff80e6ba05 (always the same)
stack pointer = 0x28:0xffffff80000a5ab0 (always the same)
frame pointer = 0x28:0xffffff80000a5b00 (always the same)
code segment = base 0x0, limit 0xfffff, type 0x1b
	     = DPL 0, pres 1, def32 1, gran 1
processor eflags = interrupt enabled, resume, IOPL = 0
current proecess = 12 (irq256: siis0)
trap number = 12
panic: page fault
cpuid = 0
uptime 1h36m3s

The crash screen on the original i386 server was basically the same.
Now that Ive moved the disks to the new server I can use this for  
whatever testing is necessary (the original server is running  
production services).

My boot/loader.conf is:
siis_load="YES"
vfs.zfs.prefetch_disable=0

Any help much appreciated,

thanks Andy.