Emulating a CPU in C++ #18 (6502) - Conditional Branches

In this episode i get the rest of the conditional branches working in the 6502 emulator.
Links:
Source for this project: github.com/davepoo/6502Emulator
Google Test: github.com/google/googletest
6502 Information: www.obelisk.me.uk/6502/
Timestamps:
0:00 Intro
1:10 BNE Test
3:16 Refactoring branch logic
5:50 BNE Implemented
6:20 BCS (Branch on Carry Set)
7:57 BCC (Branch on Carry Clear)
8:48 BMI & BPL (Branch Negative Set/Clear)
10:20 BVS & BVC (Branch Overflow Set/Clear)
12:22 Branching completed
12:50 Tidy up
14:00 Closing comments

[@CZghost]

6502 seems to have best branching instruction set, because it has branching based on all of the flags (maybe except the interrupt flag, possibly also overflow flag). But it has single addressing mode, which only seems to count with relative branching using a single byte, as it kind of limits the program to jump only 127 addresses forward and 128 addresses backwards (since it does take a sign into consideration). I would expect it to work like conditional jump, only with the flags check. So programmer has to keep track on if the code he's writing is actually compilable, because in a large scope, it could easily happen that the conditional jump would be more than a signed byte allows and it results in uncompilable code.

[@CZghost]

Okay, so overflow does have its own branch. The interrupt doesn't.

[@andrsam3682]

Are you sure that in case if branch instruction leads to other page it consumes 5 cycles, but not 4? In my opinion the doc says "2 cycles +1 if branch succeds OR +2 if to a new page", so 2 or 3 or 4.

[@DavePoo]

Yes, I think you could be right about that. Page at masswerk.at seems to confirm what you are saying. "add 1 to cycles if branch occurs on same page, add 2 to cycles if branch occurs to different page". This might also make sense if when it does the page cross, all it has to do is an extra fixup on a single byte in the PC to change the high byte of the page. I'm at the testing & debugging phase of this emulator right now so i'll address that (no pun intented), in a future part in the series.