Add async gpu wait interface (#1736)

Add async gpu wait interface
Closes #1708

src/lv_core/lv_refr.c

@@ -741,6 +741,8 @@ static void lv_refr_vdb_flush(void)
 
     /*Flush the rendered content to the display*/
     lv_disp_t * disp = _lv_refr_get_disp_refreshing();
+    if (disp->driver.gpu_wait_cb) disp->driver.gpu_wait_cb(&disp->driver);
+
     if(disp->driver.flush_cb) disp->driver.flush_cb(&disp->driver, &vdb->area, vdb->buf_act);
 
     if(vdb->buf1 && vdb->buf2) {

src/lv_draw/lv_draw_blend.c

@@ -138,7 +138,8 @@ LV_ATTRIBUTE_FAST_MEM void _lv_blend_fill(const lv_area_t * clip_area, const lv_
     const lv_area_t * disp_area = &vdb->area;
     lv_color_t * disp_buf = vdb->buf_act;
 
-
+    if (disp->driver.gpu_wait_cb) disp->driver.gpu_wait_cb(&disp->driver);
+    
     /* Get clipped fill area which is the real draw area.
      * It is always the same or inside `fill_area` */
     lv_area_t draw_area;
@@ -212,6 +213,8 @@ LV_ATTRIBUTE_FAST_MEM void _lv_blend_map(const lv_area_t * clip_area, const lv_a
     const lv_area_t * disp_area = &vdb->area;
     lv_color_t * disp_buf = vdb->buf_act;
 
+    if (disp->driver.gpu_wait_cb) disp->driver.gpu_wait_cb(&disp->driver);
+    
     /* Now `draw_area` has absolute coordinates.
      * Make it relative to `disp_area` to simplify draw to `disp_buf`*/
     draw_area.x1 -= disp_area->x1;

src/lv_hal/lv_hal_disp.h

@@ -114,7 +114,11 @@ typedef struct _disp_drv_t {
     /** OPTIONAL: Called when lvgl needs any CPU cache that affects rendering to be cleaned */
     void (*clean_dcache_cb)(struct _disp_drv_t * disp_drv);
 
+    /** OPTIONAL: called to wait while the gpu is working */
+    void (*gpu_wait_cb)(struct _disp_drv_t * disp_drv);
+
 #if LV_USE_GPU
+
     /** OPTIONAL: Blend two memories using opacity (GPU only)*/
     void (*gpu_blend_cb)(struct _disp_drv_t * disp_drv, lv_color_t * dest, const lv_color_t * src, uint32_t length,
                          lv_opa_t opa);