tft: restore ILI9341 colors and reversed logic

[tpruvot]

ILI9341 init code was recently changed to 9488 init (last 10 days)

This restore the right colors and the reversed display, like it was originally.

also, the X & Y offsets of the dogm UI are no more centered on 320x240 displays.... the dogm ui is touching the right border...

[thisiskeithb]

Maybe two similar displays need different settings? #18877 Also, @rhapsodyv has been doing a lot of work for the MKS Robin Nano + tethered TFTs.

[tpruvot]

was changed by #18901 for the colors issue (black is white now)

9341 is a 320x240, 9488 a 480x320

[thisiskeithb]

This may need some adjustment/ways to change colors for other TFTs since the colors are correct on MKS Robin Nano boards with their tethered TFTs using the previous code (#18901 (comment)).

[tpruvot]

yes, but MKS Robin nano is not using an ILI9341

[tpruvot]

1 = reverse for the 0x36 command... well... we have 3 kinds of touchscreen orientation and the define reverse our touch settings... we have a calibration app... and hmm, its not really documented for MY/MX bits... 0xE8 vs 0x28

[rhapsodyv]

The code is different, but they were in the same switch case.
Nano and tronxy is 0x8066.
Testing the right value for nano, I edited the code and forgot to separate.
@tpruvot is right and his change is welcomed, as he separate the case.

[rhapsodyv]

@tpruvot if you don't mind and have some time, take a look and test this PR in your 320x240 screen: #19017

[thinkyhead]

Funny, because when first submitted, I was going to fill in all those device init cases with what we've ended up with, but I didn't want to presume….

[rhapsodyv]

Funny, because when first submitted, I was going to fill in all those device init cases with what we've ended up with, but I didn't want to presume….

It's another think that is repeated.
After the tft io refactoring, we can try merging the init codes from emulated dogm and those from the new color ui.
It will allow more tft using both ui.

[thinkyhead]

I figured there would be some redundancy with TFT code arriving from so many quarters at the same time, but that's how we roll. Elaboration and piling up code while developing and testing, then revision and optimization once it works reliably — and once we have a reasonable understanding of how it all works.

[rhapsodyv]

Sure. I began the tft io refactoring after fixing the fsmc psram bug in 3 different places with the same code heheh.
In my other projects, I follow the same philosophy.