[Intel-gfx] [PATCH] intel: Mark cached bo as purgeable

[Shaohua Li]

On Sun, Jun 21, 2009 at 05:25:20PM +0800, Chris Wilson wrote:
> On Fri, 2009-06-19 at 09:50 +0800, Shaohua Li wrote:
> > we need to unmap the object like my reference patch. Truncate page will
> > unmap the object. And if libdrm thinks the object->virtual is still valid,
> > then things will get wrong.
> 
> I was thinking a bit more about this... We don't need to unmap the bo
> upon entering the cache as we check when we retrieve the bo from the
> cache as to whether it was truncated in the meantime. If it was
> truncated, we free the dead bo and create a new one. (Since that is more
> convenient, though we could just reset the bo.)
Yes, looks better.