06-HAL库硬件SPI DMA驱动LCD并移植LVGL 8.3

好玩的MATLAB

2ux5b5bjimq640331105.gif

关注、星标公众号，直达精彩内容
1、本节内容介绍

1.1、HAL库硬件SPI DMA在cubemx中的配置及注意事项;

1.2、HAL库SPI DMA详解与结构介绍;

1.3、使用SPI DMA驱动LCD显示屏并移植LVGL V8.3源码地址：https://gitee.com/MR_Wyf/hal-cubemx-rt-thread/tree/hal_rttNano_st7789_menu/
或者关注公众号，后台回复“SPI DMA”，获取本章节源码
2、HAL库SPI DMA在CUBEMX中的配置2.1、配置界面配置非常简单，只需要选择SPI1的TX配置为DMA模式即可，选择正常模式即可，不需要循环模式，否则LVGL可能会显示异常

3vrbxdlokq1640331205.png

2.2、SPI DMA代码详解先来看上面配置生成的代码,主要就是DMA的模式配置参数
    /* SPI1 DMA Init */
    /* SPI1_TX Init */
    hdma_spi1_tx.Instance = DMA1_Channel3;
    hdma_spi1_tx.Init.Request = DMA_REQUEST_1;
    hdma_spi1_tx.Init.Direction = DMA_MEMORY_TO_PERIPH;
    hdma_spi1_tx.Init.PeriphInc = DMA_PINC_DISABLE;
    hdma_spi1_tx.Init.MemInc = DMA_MINC_ENABLE;
    hdma_spi1_tx.Init.PeriphDataAlignment = DMA_PDATAALIGN_BYTE;
    hdma_spi1_tx.Init.MemDataAlignment = DMA_MDATAALIGN_BYTE;
    hdma_spi1_tx.Init.Mode = DMA_NORMAL;
    hdma_spi1_tx.Init.Priority = DMA_PRIORITY_LOW;
    if (HAL_DMA_Init(&hdma_spi1_tx) != HAL_OK)
    {
      Error_Handler();
    }
    __HAL_LINKDMA(spiHandle,hdmatx,hdma_spi1_tx);
继续来看下SPI DMA的接口：
HAL_StatusTypeDef HAL_SPI_Transmit_DMA(SPI_HandleTypeDef *hspi, uint8_t *pData, uint16_t Size);
HAL_StatusTypeDef HAL_SPI_Receive_DMA(SPI_HandleTypeDef *hspi, uint8_t *pData, uint16_t Size);
HAL_StatusTypeDef HAL_SPI_TransmitReceive_DMA(SPI_HandleTypeDef *hspi, uint8_t *pTxData, uint8_t *pRxData,
                                              uint16_t Size);
HAL_StatusTypeDef HAL_SPI_DMAPause(SPI_HandleTypeDef *hspi);
HAL_StatusTypeDef HAL_SPI_DMAResume(SPI_HandleTypeDef *hspi);
HAL_StatusTypeDef HAL_SPI_DMAStop(SPI_HandleTypeDef *hspi);
主要就是发送、接收，以及接收暂停、接受恢复、接收停止功能函数，本次驱动LCD主要用到的是DMA发送函数，主要有3个参数
/**
  * @brief  Transmit an amount of data in non-blocking mode with DMA.
  * @param  hspi pointer to a SPI_HandleTypeDef structure that contains
  *               the configuration information for SPI module.
  * @param  pData pointer to data buffer
  * @param  Size amount of data to be sent
  * @retval HAL status
  */
HAL_StatusTypeDef HAL_SPI_Transmit_DMA(SPI_HandleTypeDef *hspi, uint8_t *pData, uint16_t Size);
2.3 LCD驱动函数改造上一章节中我们使用的是SPI驱动的LCD，并没有加入DMA，本章节咱们对驱动函数进行改造，加入DMA驱动,只需要把我们的SPI发送函数改为DMA发送函数即可，以下几个函数同理改动
// ST7789写函数
static HAL_StatusTypeDef lcd_st7789_write(int is_cmd, uint8_t data)
{
uint8_t pData[2] = { 0 };
assert_param(NULL != hspi_lcd);
pData[0] = data;
if (is_cmd)
  HAL_GPIO_WritePin(LCD_DC_GPIO_Port, LCD_DC_Pin, GPIO_PIN_RESET);
else
  HAL_GPIO_WritePin(LCD_DC_GPIO_Port, LCD_DC_Pin, GPIO_PIN_SET);
// return HAL_SPI_Transmit(hspi_lcd, pData, 1, HAL_MAX_DELAY);
  return HAL_SPI_Transmit_DMA(hspi_lcd, pData, 1);
}
/********************************************************************
*
*       LcdWriteReg
*
* Function description:
*   Sets display register
*/
void lcd_st7789_write_reg(uint8_t Data)
{
HAL_GPIO_WritePin(LCD_DC_GPIO_Port, LCD_DC_Pin, GPIO_PIN_RESET);
// HAL_SPI_Transmit(&hspi1, &Data, 1, 10);
HAL_SPI_Transmit_DMA(&hspi1, &Data, 1);
}
/********************************************************************
*
*       LcdWriteData
*
* Function description:
*   Writes a value to a display register
*/
void lcd_st7789_write_data(uint8_t Data)
{
HAL_GPIO_WritePin(LCD_DC_GPIO_Port, LCD_DC_Pin, GPIO_PIN_SET);
HAL_SPI_Transmit_DMA(&hspi1, &Data, 1);
//HAL_SPI_Transmit(&hspi1, &Data, 1, 10);
}
/********************************************************************
*
*       lcd_st7789_write_data_multiple
*
* Function description:
*   Writes multiple values to a display register.
*/
extern uint8_t g_spi_dma_tc;
void lcd_st7789_write_data_multiple(uint8_t *pData, int NumItems)
{
if (g_spi_dma_tc) {
  g_spi_dma_tc = 0;
  HAL_GPIO_WritePin(LCD_DC_GPIO_Port, LCD_DC_Pin, GPIO_PIN_SET);
  //HAL_SPI_Transmit(&hspi1, pData, NumItems, 10);
  HAL_SPI_Transmit_DMA(&hspi1, pData, NumItems);
}
}
3、移植LVGL V8.3本次移植的是lvgl V8.3,源码在lvgl官方github上就可以下载到，想下载的兄弟小手动一动，不想下载的也没关系，小飞哥会把源码开源，直接拿过去就行了
至于LVGL的移植，就不再赘述了，相信网上有成堆的教程，小飞哥也不再废话浪费大家伙时间了，直接下载源码即可
主要强调几个移植的点：

1、周期调用lvgl tick接口，提供lvgl“心跳”在定时器3回调函数中调用lv_tick_inc(1)，为LVGL提供心跳，周期10ms足够
void HAL_TIM_PeriodElapsedCallback(TIM_HandleTypeDef *htim)
{
  /**timer for lvgl,period 1ms*/
  if (htim->Instance == TIM3)
  {
    lv_tick_inc(1);
  }
  if (htim->Instance == TIM15)
  {
    // if(RT_EOK==rt_sem_take(sem_uart_rec,RT_WAITING_NO))
    // {
    if (embedded_get_uart_rec_flag())
    {
      /*100ms*/
      if (embedded_get_uart_timeout_cnt() > 9)
      {
        //      lv_tick_inc(1);
        embedded_set_uart_rec_flag(RT_FALSE);
        rt_sem_release(sem_uart_timeout);
      }
    }
    // }
  }
}

2、lvgl初始化配置，使用“双缓存”void lv_port_disp_init(void)
{
    /*-------------------------
     * Initialize your display
     * -----------------------*/
    disp_init();
    /*-----------------------------
     * Create a buffer for drawing
     *----------------------------*/
    /**
     * LVGL requires a buffer where it internally draws the widgets.
     * Later this buffer will passed to your display driver's `flush_cb` to copy its content to your display.
     * The buffer has to be greater than 1 display row
     *
     * There are 3 buffering configurations:
     * 1. Create ONE buffer:
     *      LVGL will draw the display's content here and writes it to your display
     *
     * 2. Create TWO buffer:
     *      LVGL will draw the display's content to a buffer and writes it your display.
     *      You should use DMA to write the buffer's content to the display.
     *      It will enable LVGL to draw the next part of the screen to the other buffer while
     *      the data is being sent form the first buffer. It makes rendering and flushing parallel.
     *
     * 3. Double buffering
     *      Set 2 screens sized buffers and set disp_drv.full_refresh = 1.
     *      This way LVGL will always provide the whole rendered screen in `flush_cb`
     *      and you only need to change the frame buffer's address.
     */
    /* Example for 1) */
    static lv_disp_draw_buf_t draw_buf_dsc_1;
    static lv_color_t buf_1[MY_DISP_HOR_RES * 10];                             /*A buffer for 10 rows*/
    lv_disp_draw_buf_init(&draw_buf_dsc_1, buf_1, NULL, MY_DISP_HOR_RES * 10); /*Initialize the display buffer*/
    /* Example for 2) */
    static lv_disp_draw_buf_t draw_buf_dsc_2;
    static lv_color_t buf_2_1[MY_DISP_HOR_RES * 10];                                /*A buffer for 10 rows*/
    static lv_color_t buf_2_2[MY_DISP_HOR_RES * 10];                                /*An other buffer for 10 rows*/
    lv_disp_draw_buf_init(&draw_buf_dsc_2, buf_2_1, buf_2_2, MY_DISP_HOR_RES * 10); /*Initialize the display buffer*/
    /* Example for 3) also set disp_drv.full_refresh = 1 below*/
    static lv_disp_draw_buf_t draw_buf_dsc_3;
    static lv_color_t buf_3_1[MY_DISP_HOR_RES * MY_DISP_VER_RES]; /*A screen sized buffer*/
    static lv_color_t buf_3_2[MY_DISP_HOR_RES * MY_DISP_VER_RES]; /*Another screen sized buffer*/
    lv_disp_draw_buf_init(&draw_buf_dsc_3, buf_3_1, buf_3_2,
                          MY_DISP_VER_RES * MY_DISP_VER_RES); /*Initialize the display buffer*/
    /*-----------------------------------
     * Register the display in LVGL
     *----------------------------------*/
    static lv_disp_drv_t disp_drv; /*Descriptor of a display driver*/
    lv_disp_drv_init(&disp_drv);   /*Basic initialization*/
    /*Set up the functions to access to your display*/
    /*Set the resolution of the display*/
    disp_drv.hor_res = MY_DISP_HOR_RES;
    disp_drv.ver_res = MY_DISP_VER_RES;
    /*Used to copy the buffer's content to the display*/
    disp_drv.flush_cb = disp_flush;
    /*Set a display buffer*/
    disp_drv.draw_buf = &draw_buf_dsc_2;
    disp_drv_p = &disp_drv;
    /*Required for Example 3)*/
    // disp_drv.full_refresh = 1;
    /* Fill a memory array with a color if you have GPU.
     * Note that, in lv_conf.h you can enable GPUs that has built-in support in LVGL.
     * But if you have a different GPU you can use with this callback.*/
    // disp_drv.gpu_fill_cb = gpu_fill;
    /*Finally register the driver*/
    lv_disp_drv_register(&disp_drv);
}
刷新函数：
/*Flush the content of the internal buffer the specific area on the display
*You can use DMA or any hardware acceleration to do this operation in the background but
*'lv_disp_flush_ready()' has to be called when finished.*/
static void disp_flush(lv_disp_drv_t *disp_drv, const lv_area_t *area, lv_color_t *color_p)
{
    if (disp_flush_enabled)
    {
        // /*The most simple case (but also the slowest) to put all pixels to the screen one-by-one*/
        //        int32_t x;
        //        int32_t y;
        //        for (y = area->y1; y y2; y++)
        //        {
        //            for (x = area->x1; x x2; x++)
        //            {
        //                /*Put a pixel to the display. For example:*/
        //                /*put_px(x, y, *color_p)*/
        //                lcd_st7789_write_pixel(x, y, color_p->full);
        //                color_p++;
        //            }
        //        }
        // int32_t y;
        // lcd_st7789_set_addr_win(area->x1, area->y1, area->x2, area->y2); // 指定填充区域
        //                                                                  // 一行一行 DMA
        // for (y = area->y1; y y2; y++)
        // {
        //     lcd_st7789_write_data_multiple((uint8_t *)color_p, (uint16_t)(area->x2 - area->x1 + 1) * 2);
        //     color_p += (area->x2 - area->x1 + 1);
        // }
        unsigned int size = (area->x2 - area->x1 + 1) * (area->y2 - area->y1 + 1) * 2;
        lcd_st7789_set_addr_win(area->x1, area->y1, area->x2, area->y2); // 指定填充区域
        lcd_st7789_write_data_multiple((uint8_t *)color_p, size);
    }
    /*IMPORTANT!!!
     *Inform the graphics library that you are ready with the flushing*/
    //    lv_disp_flush_ready(disp_drv);
}

3、移植后的目录如下