Around 11 o'clock on Nov 15, Branden Robinson wrote:
> As I understand it, "packed pixel modes" are those in which pixel data > is not aligned on (Intel 80386+) word boundaries, which are 32 bits. That's not the historical usage. In the old days, there were two kinds of frame buffers -- packed pixel and planar. Planar frame buffers had multiple 1-bit frame buffers that were joined together in hardware to form wider pixels. The Apollo workstations used such. It can make writing graphics code a bit simpler as you can use the monochrome code for everything and just loop over the list of bit planes. Packed pixel meant pixels that were stored together in the same bytes of memory. Pretty much everyone agrees that planar frame buffer suck, so you don't see them anymore often than you find 36 bit CPUs. > IIRC packed pixel modes were a little bit of a hack introduced by some > video card manufacturers (like ATI) so that you could squeeze bigger > 2D framebuffers into smaller amounts of RAM, since at depth 24 but 32 > bits per pixel, 8 bits of information were unused. While saving memory was one reason, a far more interesting one is performance. Packing pixels together like this means that you get an automatic 25% savings in memory bandwidth needed to shuffle pixels around the frame buffer. Of course, it sucks for the CPU, but we're pretty much used to dealing with brain damaged graphics hardware in the PC world. [EMAIL PROTECTED] XFree86 Core Team SuSE, Inc. _______________________________________________ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert
