How to allocate more memory for the Framebuffer - STM32H7

How-to
1

Description

So I’ve been dealing with this issue for awhile now and need some serious guidance on how to resolve it. I’m currently running LVGL on a Alientek Dev board utilizing an STM32H7 to drive a 1020x600 px display but cannot allocate enough memory to create an array large enough to actually fill the entire screen. At best I can render maybe 1020x200px which does indeed display the LVGL demo I have running, so LVGL’s APIs appear to be working as expected.

What MCU/Processor/Board and compiler are you using?

STM32H7 with the most recent version of LVGL.

What do you want to achieve?

LVGL is indeed running and working, I just need direction on how to properly allocate memory and or write/update the flash so I can utilize an array large enough to actually fill the entire screen.

What have you tried so far?

Using a secondary array allocated to a separate memory address to write. This sort of worked but I was getting an issue with the alpha were I couldn’t contain the layer to a set section of the screen so the layer behind it wasn’t visible unless I set the alpha way down which makes the image dark and difficult to read.

Attempting to setup DMA and use the SDRAM I know is one the board, configuring the FMC did not work. I also confirmed that the hardware is consistent as it matches the pinout provided in my documentation.

Code to reproduce

This is all I am currently doing to update the image.

 *(__IO uint16_t*) (hltdc.LayerCfg[0].FBStartAdress + (2*(Ypos*1040 + Xpos))) = (uint16_t)RGB_Code;

******************************************************************************
 * @attention
 *
 * <h2><center>&copy; Copyright (c) 2020 STMicroelectronics.
 * All rights reserved.</center></h2>
 *
 * This software component is licensed by ST under BSD 3-Clause license,
 * the "License"; You may not use this file except in compliance with the
 * License. You may obtain a copy of the License at:
 *                        opensource.org/licenses/BSD-3-Clause
 *
 ******************************************************************************
 */

/* Entry Point */
ENTRY(Reset_Handler)

/* Highest address of the user mode stack */
_estack = ORIGIN(RAM_D1) + LENGTH(RAM_D1);	/* end of "RAM_D1" Ram type memory */

_Min_Heap_Size = 0x200 ;	/* required amount of heap  */
_Min_Stack_Size = 0x400 ;	/* required amount of stack */


/* Memories definition */
MEMORY
{
  DTCMRAM    (xrw)    : ORIGIN = 0x20000000,   LENGTH = 128K
  ITCMRAM    (xrw)    : ORIGIN = 0x00000000,   LENGTH = 64K
  RAM_D1     (xrw)    : ORIGIN = 0x24000000,   LENGTH = 512K
  RAM_D2     (xrw)    : ORIGIN = 0x30000000,   LENGTH = 288K
  RAM_D3     (xrw)    : ORIGIN = 0x38000000,   LENGTH = 64K
  FLASH      (rx)     : ORIGIN = 0x08000000,   LENGTH = 1408K
  DATA       (xrw)    : ORIGIN = 0x080E0000,   LENGTH = 640K
}
```c
2

1020x600x2bpp means you have no choice but to use SDRAM, as that requires 1.2MB for a framebuffer. Are you using the BSP drivers ST provides?

3

Not that I’m aware of. I’m not even sure which functions to start with in that regard.

I’ve configured FMC through the CubdeIDE and then created a memory region starting at 0xC0000000 where I attempted to write the framebuffer to but that did not work. I’m very lost in terms of how to actually get SDRAM working/implemented as their is not a ton of resources for it so you can assume I’m starting with nothing.

4

I have a STM32H743 board from waveshare. They provide some code for setting up the SDRAM.
It’s under BSP_SDRAM_Init. I would suppose that Alientek provides some equivalent examples.

I have setup the SDRAM on address 0xd0000000. I think that depends on configuration in SDRAM init function.

When the FMC is properly setup for SDRAM you have to configure (adjust) the LTCD (MX_LTDC_Init or something like that).

And at last, adjust the flush function to write the lvgl created data to your new frame buffer (in SDRAM).

5

Thanks for the suggestion I’ll have a look.