readme.md

@@ -14,18 +14,15 @@ Non-goals:
 
 Enjoy! I'll probably work on this off and on for the next few weeks until I've got it producing fractals.
 
--- brooke, january 2023 - december 2024
+-- brooke, january 2023 - february 2024
 
 ## Current state
 
-Basic rendering is functional, with interactive zoom/pan (+/-/arrows) and 4 preset viewports via the number keys.
+Basic rendering is functional, but no interactive behavior (zoom/pan) or benchmarking is done yet.
 
-The 16-bit signed integer multiplication takes two 16-bit inputs and emits one 32-bit output in the zero page, using the Atari OS ROM's floating point registers as workspaces. Inputs are clobbered.
+The 16-bit signed integer multiplication works; it takes two 16-bit inputs and emits one 32-bit output in the zero page, using the Atari OS ROM's floating point registers as workspaces. Inputs are clobbered.
 
-* 16-bit multiplies are decomposed into 4 8-bit unsigned multiplies and some addition
+The main loop is a basic add-and-shift, using 16-bit adds which requires flipping the sign of negative inputs (otherwise you'd have to add all those sign-extension bits). Runs in 470-780 cycles depending on input.
-* an optimized case for squares uses a table of 8-bit squares to reduce the number of 8-bit multiplication sub-ops
-* when expanded RAM is available as on 130XE, a 64KB 8-bit multiplication table accelerates the remaining multiplications
-* without expanded RAM, a table of half-squares is used to implement the algorithm from https://everything2.com/title/Fast+6502+multiplication
 
 The mandelbrot calculations are done using 4.12-precision fixed point numbers. It may be possible to squish this down to 3.13.
 
@@ -33,18 +30,17 @@ Iterations are capped at 255.
 
 The pixels are run in a progressive layout to get the basic shape on screen faster.
 
-There is a running counter of ms/px using the vertical blank interrupts as a timer, used to track our progress. :D
+## Next steps
 
-There's a check for cycles in (zx,zy) output when in the 'lake'; if values repeat, they cannot escape. This is a big time saver in fractint.
+Add a running counter of ms/px using the vertical blank interrupts as a timer. This'll show how further work improves it!
 
-There's some cute color cycling.
+Check for cycles in (zx,zy) output when in the 'lake'; if values repeat, they cannot escape. This is a big time saver in fractint.
+
+I may be able to do a faster multiply using tables of squares for 8-bit component multiplication.
+(done)
 
 ## Deps and build instructions
 
 I'm using `ca65` as a macro assembler, and have a Unix-style `Makefile` for building. Should work fairly easily on Linux and Mac. Might work on "raw" Windows but I use WSL for that.
 
 Currently produces a `.xex` executable, which can be booted up in common Atari emulators and some i/o devices.
-
-## Todo
-
-See ideas in `todo.md`.

todo.md

@@ -2,13 +2,14 @@ things to try:
 
 * patch the entire expanded-ram imul8xe on top of imul8 to avoid the 3-cycle thunk penalty :D
 
+* optimize out a store/load with mul8_add16 and mul8_add24
+
 * try 3.13 fixed point instead of 4.12 for more precision
   * can we get away without the extra bit?
 
 * y-axis mirror optimization
 
 * 'wide pixels' 2x and 4x for a fuller initial image in the tiered rendering
-  * maybe redo tiering to just 4x4, 2x2, 1x1?
 
 * extract viewport for display & re-input via keyboard