When identifying a selected `00000001.sav`, cmac type testing is first attempted for 3DS saves with the **DISA** format. If a 3DS save data is not detected, the search continues at expected offsets for gba vc bottom slots in increasingly larger save sizes. For the GBA VC saves, this only calculates and corrects whichever of the two slots is newer or has the higher counter. There are five additional **CMAC_AGBSAVE_** types to differentiate the different gba save sizes. If a bottom slot was found but determined to be older based on comparing the counter values, the search defaults to **CMAC_AGBSAVE_SD**.
For the save counters, there are two scenarios where the value `00` is considered bigger than `FF` depending which slot has what value. If clarification is required, I will post a diagram at [issue #494](https://github.com/d0k3/GodMode9/issues/494) explaining what is meant by `00` bigger than `FF`.
Please make corrections and improvements to the coding as you see fit in your review. I am not well versed or efficient in the C/C++ programming language.
This is meant to replace the Prompt I was using previously.
Fun fact: WarioWare DIY seems to have *something* on the SPI bus, as it
returns an ID of 0x000001 consistently. Or am I just glitching the
parallel flash? Or did I get a fake?
* Fixed card reading (I didn't notice I broke it)
* Fixed cart swapping
* Still unsupported: CTR carts
* Still hangs if cart has no SPI flash (see WarioWare DIY)
The initialization code was simplified to only enable the SPICARD
interface, which assumes the *TRCARD interface has been initialized
beforehand.
To keep things simple, I just wrapped SPIWriteRead with SPICARD_Init and
SPICARD_deinit (which does the exact opposite)
For some reason (messed up memory access?) the first two time I read the
status register, I get garbage.
Also:
* Reinserting card breaks SPI (everything reads 0xff
* No support for CTR carts for now
