Remove the insane infrastructure for preallocating coheren DMA regions,
and just allocate the memory where needed.  This also gets rid of the
aligment adjustments given that Documentation/DMA-API-HOWTO.txt sais:

  "The cpu return address and the DMA bus master address are both
   guaranteed to be aligned to the smallest PAGE_SIZE order which
   is greater than or equal to the requested size.  This invariant
   exists (for example) to guarantee that if you allocate a chunk
   which is smaller than or equal to 64 kilobytes, the extent of the
   buffer you receive will not cross a 64K boundary."
Signed-off-by: NChristoph Hellwig <hch@lst.de>
[djbw: moved allocation from start to init, re-add memset]
Signed-off-by: NDan Williams <dan.j.williams@intel.com>