>There's another way? :-/ >My reading of the drivers so far made it seem like an agp driver HAS to do >all of: > >1. allocate actual RAM when an ioctl(ALLOC) is done >2. "bind" it by making an entry in the GATT when ioctl(BIND) is done >3. mmap the RAM in step #1 to userspace when an mmap() is requested. > >But you seem to imply there's a more direct method possible?
Some chipsets (and the original agpgart supported those only) can let the CPU access the AGP aperture directly. All mmap had to do was then to map the user pages to the aperture physical pages, if they had memory bound or not didn't matter. Ben. _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
