[Oberon] System-V - Frame Clipping
joerg.straube at iaeth.ch
Mon Mar 4 07:36:30 CET 2019
Clipping in HW is a looong way to go.
Remember: Currently the interface to the display HW is a plain memory access to the video framebuffer.
All primitives like Dot, ReplConst and so on implemented in Display.Mod (and Line, Circle, Elipses implemented in Display1.Mod not included by default in PO13) are written in Oberon.
Before implementing clipping in HW you would have to implement these primitives in HW. After having done that, then you can think of clipping those primitives.
Basically you have to implement a GPU in Verilog...
> Am 04.03.2019 um 06:32 schrieb Skulski, Wojciech <skulski at pas.rochester.edu>:
> the textbooks by Pong P. Chu discuss the video controller implementation which is doing image blending in hardware (firmware, actually). The idea is close to what you are suggesting, I think. Both these books, one in VHDL and the other in Verilog, are very well written and worth reading. I believe that good inspiration on clipping in hardware can come from reading these chapters.
> Go to RiskFive.com and select the "Web Resources" section. The first two "Books and Articles" are the ones I am referring to.
> From: Oberon [oberon-bounces at lists.inf.ethz.ch] on behalf of Tomas Kral [thomas.kral at email.cz]
> Sent: Sunday, March 3, 2019 1:26 PM
> To: oberon at lists.inf.ethz.ch
> Subject: Re: [Oberon] System-V - Frame Clipping
>> VAR F: Display.Frame
>> Clip(F) (*clipping ON*)
>> Clip(NIL) (*clipping OFF*)
> RISC-5 has some reserved address space from $FFFF00 to $FFFFFF, some
> used by H/W mapped registers for I/O.
> I was thinking, to use some of those for `VID.v' clipping.
> Procedure `Clip()' would fiddle with some of these reserved addresses.
> Is it worthwhile, or
> just bad idea, difficult, complex or costing many cycles, and
> stalling CPU?
> Tomas Kral <thomas.kral at email.cz>
> Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems
More information about the Oberon