This paves the way for accelerants to map the framebuffer in
user memory, not kernel memory. It uses the ServerMemoryAllocator
in order to avoid mapping the framebuffer multiple times in
client applications.
While at it, clean up some names.
Tested with DirectWindowStars and GLTeapot, both work.
Change-Id: I4e0c003fae99891044af0a5e6e1d6a4506da6430
Reviewed-on: https://review.haiku-os.org/c/haiku/+/10571
Tested-by: Commit checker robot <[email protected]>
Reviewed-by: waddlesplash <[email protected]>
Augustin Cavalier
f6be811473