Re: [RFQ] New driver architecture questions

[Matthew Wilcox <matthew@wil.cx>]

On Fri, May 15, 2009 at 12:03:39PM -0600, Mukker, Atul wrote:
> > The solution to "We have some people who speak French and other people who
> > speak German" is not to invent Esperanto ;-)
> [Atul] We really wish they could communicate in English :-), since that's not an option, we agree in principle that using native Linux Kernel APIs wherever possible is probably a good idea.

I'd stick to the C APIs where possible ... oh, that's what Linux does.  OK ;-)

> > Using one or the other internally is fine (we don't care what you do),
> > but we want to see memcpy().  By the way, the documentation I found for
> > ScsiPortMoveMemory() seems to indicate that it's memmove(), not memcpy().
> > Mapping memcpy() to ScsiPortMoveMemory() is fine ... but you can't
> > realiably go the other way.
> [Atul] It's actually memcpy(),http://msdn.microsoft.com/en-us/library/ms805434.aspx

No, it's memmove().  "The (ReadBuffer + Length) can overlap the area
pointed to by WriteBuffer."

-- 
Matthew Wilcox				Intel Open Source Technology Centre
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours.  We can't possibly take such
a retrograde step."