marcmerlin / framebuffer_gfx Goto Github PK

In the Framebuffer_GFX.cpp line 152:

return (uint32_t)(c.r<<16) + (uint32_t)(c.g<<8) + (uint32_t)c.b;

you used incorrect cast to uint32_t

The expression

(uint32_t)(c.r<<16)

means that you make left shift FIRST, compute the result, using the type of the c.r variable and only after that (and perhaps after truncating the result to uint16 on AVR controllers) - casting it to uint32_t.

The correct syntax that prevent the truncation is:

((uint32_t)c.r<<16)

so you make casting FIRST and than shift the uint32_t value.

Hi!

I'm trying to port awtrix-light (https://github.com/Blueforcer/awtrix-light) to HUB75 panels, using https://github.com/mrfaptastic/ESP32-HUB75-MatrixPanel-DMA.

Awtrix-light uses FastLED_NeoMatrix, which in turn uses Framebuffer_GFX.

Does it make sense I add support for ESP32-HUB75-MatrixPanel-DMA to Framebuffer_GFX?

Thank you!

Board: ESP32
Platform: PlatformIO

Using SmartMatrix. (or trying to use)

I have tried filling the screen with a color using:

#define LED_RED_HIGH 		  (31 << 11)
matrix->drawRect(0, 0, kMatrixWidth, kMatrixHeight, LED_RED_HIGH);
matrix->fillRect(0, 0, kMatrixWidth, kMatrixHeight, LED_RED_HIGH);
// or
matrix->fillScreen(LED_RED_HIGH);

and in both cases I get:

Not Implemented in Framebuffer::GFX

How are these not implemented ??? How do I fill the screen with a color without using backgroundLayer.fillScreen() ?