The Correctness PR

[rynorris]

Finally after many hours I tracked down the bug causing mario to get stuck at x=0 in SMB. In the process of doing so I added:

• Ability to hold Tab key while playing to record CPU trace. (reduces the NES to about 1/100 speed)
• Framework for tests which compare PPU output to saved BMPs
• Framework for tests which run blargg's test ROMs.

It turned out the bug was a subtlety in my SBC logic where in one very particular situation the carry flag was set incorrectly.

As a result:

• We now pass all the official instruction tests in instr_test-v5, and have a regression test for it.
• SMB glitch is fixed. The game seems pretty flawless now, but something is still slightly off with X movement. The demo at the start does not play properly. But it's not really noticeable in-game.
• Metroid bugginess is fixed. No obvious glitches anymore.
• Megaman2 menu screen is fixed. Before it was very corrupted and you couldn't get in-game. Now I just played a whole stage with no obvious bugs!

These issues have been plaguing me for days, to realize they were all caused by the same tiny bug is crazy!

[codecov-io]

Codecov Report

Merging #12 into master will not change coverage. The diff coverage is 100%.

@@           Coverage Diff           @@
##           master      #12   +/-   ##
=======================================
  Coverage   93.32%   93.32%           
=======================================
  Files          38       38           
  Lines        4916     4916           
=======================================
  Hits         4588     4588           
  Misses        328      328

Impacted Files	Coverage Δ
src/emulator/cpu/instructions.rs	100% <100%> (ø)	:arrow_up:
src/emulator/test/mod.rs	94.73% <100%> (ø)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update b2b6e34...c1b5c75. Read the comment docs.