nesdev.org

lidnariq wrote:With ld65, you really have to structure your LoROM application as a series of 32 KiB slices. Look at tepples's lorom template

Very helpful, thanks!

lidnariq wrote: The /ROMSEL signal on the card edge is literally just "low if ((A22 high or A15 high) and (A23-A17 not equal to 0x7E))". That's the closest thing you get to "the SNES's "native" memory map".

That sounds interesting, where did you get that Information?

I'm confused about how the PC is loaded if I do a JML or JSL. If I understand the mapping correctly, it means that SNES Address ROM Address $00:8000 $00:0000 $01:8000 $00:8000 $02:8000 $01:0000 . . . And so on. How do absolute long jumps then work? Let's say I tell my assembler to place my reset rou...

Reset Vector (00FFDH, 00FFCH) 00fffd, you forgot an f. Also, this is actually at position 007ffc-d of the ROM, due to the way ROMs are always loaded to 8000-ffff in the memory map. (At least for LoROM) how does the SNES know how to map/mirror $00:$7FFC to $FF:$FFFC? This is my very dumb answer. One...

Thank you for your quick reply! I just found a document showing that the SNES will on reset "Jump to the address indicated by Reset Vector (00FFDH, 00FFCH)". Now off to some joypad input and movement

I'm wondering how the SNES finds the reset vector in my ROM file. I understand that the 65816 processor looks at the last 16 bytes for the reset, NMI, IRQ, etc. addresses(both for emulation and native mode). I have written a little program, that will load some sprite data into VRAM, set some registe...

Yes, I can see that now. I know, this wouldn't work on real hardware, but I thought a emulator would give me enough leeway that I could get away with some quick&dirty code(NES emulators seem to be more forgiving in that regard). let's see if I get finally get a sprite displayed tonight. NMI/VBla...

Busy weekend, but today I had some time to code again. And, of course, I ran into a problem again. Here's another weird behavior: whenever I load my ROM into bsnes emulator, at first it won't work(the Properties viewer will show that some of my initial DMA register settings are off, therefore no pro...

It makes me wonder if ca65 is confused by the lower half of the byte you're passing through rep; you don't need to set it to 8. No, that wasn't a problem, it was really switching into native mode after setting the status bits. But so far I am moving along fine, I managed to load CG RAM with my pale...

I found one of the reasons why my code wouldn't work: I set the A, X and Y registers to 16 bits ( rep #$38 ) before setting the processor to native mode ( clc followed by xce ). Therefore all following opcodes were translated wrongly(with 8 bit registers) as I found out with a help of a disassembler...

That format advice is a good one. I was really wondering why there is no "meta header" like they use for NES ROMs for emulators - which are removed if you burn it into a chip to use on a actual NES. Fun, but you need the proper tools like a ROM burner(which I luckely own).

Wow, thank you! I never used .smart before, since it is my first 65816 code. A quick look up in the ca65 doc tells me I probably need to study the registers/architecture of the 65816 better - my guess right now is, I'm not using Data Bank and Program Bank registers properly. That would also explain ...

nicklausw wrote:Another thing, your header has the title set to 18 bytes instead of 21. Fixing this made a red screen come up, not sure why that's happening but it's better than what was showing up before.

A red screen is what I want! I changed my header, but it still won't work, what emulator do you use?

Another problem I see is that you use jsr to call your init code, but that's where you set up the stack, so rts won't return you back where you were, because the stack pointer is no longer pointing at the address you need to return to. You should probably jmp to the init code, and jmp back instead....

Nicole wrote:I notice that your init code already turns off forced blanking. That might be interfering with your writes to $2122, since they're probably not happening during Vblank/Hblank.

True, I moved it to NMI/VBlank, to no avail. Still a black screen.