CPU::executeBranch may calculate wrong m_skipCycles

[Froser]

According to Opcode matrix from http://www.oxyron.de/html/opcodes02.html:

"*" : add 1 cycle if page boundary is crossed.
add 1 cycle on branches if taken.

From cpu.cpp line 349:
if (branch) { int8_t offset = m_bus.read(r_PC++); ++m_skipCycles; auto newPC = static_cast<Address>(r_PC + offset); setPageCrossed(r_PC, newPC, 2); r_PC = newPC; }

If the branch is taken, and page is crossed, m_skipCycles will increase by 3. But according to the document, only 2 cycles should be added.
Is there any mistake, or did I misunderstand the document?
Thank you.

[amhndu]

Hi!
I was mostly following this IIRC. There it says:

2 (+1 if branch succeeds
+2 if to a new page)

I think I may have misunderstood this to mean that it adds +1 if branch succeeds and another +2 if to a new page.
Another source here also seems to match what you point out. I'll run some test roms and do some more digging. Thanks for pointing out!