>On Sunday, June 8, 2003, at 09:33  PM, Henry B. Hotz wrote:
>
>> At 8:32 PM -0400 6/8/03, Michael R.Zucca wrote:
>>> 3. Do un-16-byte-aligned transfers under 4k by copying the data to/from a pre-allocated and aligned transfer buffer. This will help solve the sync negotiation problem by boiling it down to figuring out how to do odd-sized transfers that appear to be completely DMA.
>>> 4. Figure out a way to do odd-sized reads/writes that will satisfy the sync negotiation code.
>>
>> Doesn't most of the disk IO occur in 512-byte increments from page-aligned buffers?
>
>During boot, there are some transactions occur that aren't aligned or of even size, or at least they don't seem to be. This may be some poor coding in the NCR code, but it does happen.
>
>Experience seems to show that once the probing at boot is finished, most transfers are the size of a page or some multiple thereof, which I would expect.


The only data xfer operations that I would expect are the inquiries and
non-"data" stuff that happens at bus probe/scan time.  As you say,
subsequent operations should be 100% in multiples of 512.

-dgl-