CVS Update: xc (branch: trunk)

[Benjamin Herrenschmidt]

> Right now I don't expect any driver to do component alpha text.  So I
> don't see the point in adding a flag before we make exa help drivers get
> this done (and at that point, we could remove the check from the commit
> in question).  The patch I was working on would have been a small step
> towards being able to do a CA helper (it added functions to pin pixmaps,
> which we would need in order to do this), but I couldn't get it to make
> this same improvement to text performance for some reason.

Fair enough.

Note about pinned pixmaps: I really have a problem with those. I'm
already getting calls in the current X.org CVS to PrepareAccess for
pixmaps that are pinned and of a different bpp as the front buffer.

This is illegal. If I ever implement EXA on a card where I cannot
simultaneously "prepare" several pixmaps with different bpp's (for
example I don't think I have the surface stuff on r128 so I can't set
separate endian swappers), I _need_ to be able to force a migration to
memory by returning FALSE;

The way I defined PrepareAccess(), the only pinned pixmap case that is
allowed is for pixmaps that have the same bpp as the front buffer.

I should update the exa docs to reflect that limitation I suppose ...

Ben.