I wrote:

>the assembly generated (my comments) is
>
>  0:   00852021        addu    $a0,$a0,$a1     # add first two args
>  4:   2402000d        li      $v0,13          # load constant offset 13
>  8:   3c010001        lui     $at,0x1         # load $at with (1 << 16)
>  c:   00240821        addu    $at,$at,$a0     # add to frist two args
> 10:   03e00008        jr      $ra             # delayed-branch  return
> 14:   a0228020        sb      $v0,-32736($at) # store into
>                                                # (a0 + a1) + (1 << 16) +
                                                   (0x10000 - 3800)

oops, that last line has the operands backwards. it should be
			# (a0 + a1) + (1 << 16) +
			    (32800 - (1 << 16))

if it wasnt clear from context, the upper part is done in 32-bit
arithmetic, and the 16-bit immediate offset in the bottom line gets
sign-extended to 32 bits, so it all works out properly.