[U-Boot] [PATCH V2 1/3] memcpy: copy one word at a time if possible

[Peter Tyser]

On Thu, 2009-10-08 at 18:00 +0200, Alessandro Rubini wrote:
> > No interest in the suggestion to not require count to be an exact
> > multiple of 4/8?
> 
> Actually, I wrote about that in my patch 0/3.

Sorry, I should have read more thoroughly.

> > I don't think it would be that hard to update the logic accordingly
> > and this would let your code be utilized much more often, especially
> > if/when we run on a 64-bit machine.
> 
> That's true, but I think the most important case is lcd scrolling,
> where it's usually a big power of two -- that's where we had the #ifdef,
> so the problem was known, I suppose.

I think the most important case for *you* is lcd scrolling, but for 99%
of everyone else, it isn't at all:)  memcpy() and memset() are used 100
times more often in non-lcd related code and most boards don't even have
LCDs.  In my opinion, we shouldn't be fixing/bloating common code for 1
outlying scenario, we should be trying to improve it for common cases.

> Currently, I don't even know if this is going to be picked up, so I
> don't want to go too far -- and in that case I would like to measure
> things and be able to test for stupid bugs, so it takes time.

Sure, I understand where you're coming from.  FWIW, if people don't want
to pick this up as it affects lots of people, you could always add an
architecture-specific memcpy/memset implementation.

> If there's interest, there's memmov to fix; it's used pretty often and
> it's the same idea as memcpy (well, scrolling should use memmove,
> in theory).

I'll quit pestering, just wanted to put my $.02 in:)

Best,
Peter