Mirrors_UI_SDK/lvgl
1
0
Fork 0
mirror of https://github.com/lvgl/lvgl.git synced 2026-05-21 14:32:44 +08:00
Code Issues Actions 47 Packages Projects Releases Wiki Activity

feat(NemaGFX): encoded images (#8523)

Browse Source
This commit is contained in:
Liam Howatt
2025-07-28 08:28:39 -04:00
committed by GitHub
parent 8d74df6ae8
commit 0f3703e220
1 changed files with 22 additions and 73 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/draw/nema_gfx/lv_draw_nema_gfx_img.c
+22 -73
View File
@@ -41,9 +41,9 @@
* STATIC PROTOTYPES
**********************/
static void _draw_nema_gfx_tile(lv_draw_task_t * t, const lv_draw_image_dsc_t * dsc, const lv_area_t * coords);
static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * dsc, const lv_area_t * coords);
static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * dsc,
const lv_image_decoder_dsc_t * decoder_dsc, lv_draw_image_sup_t * sup,
const lv_area_t * img_coords, const lv_area_t * clipped_img_area);
static uint32_t lv_nemagfx_mask_cf_to_nema(lv_color_format_t cf);
@@ -51,80 +51,29 @@ static uint32_t lv_nemagfx_mask_cf_to_nema(lv_color_format_t cf);
* STATIC FUNCTIONS
**********************/
static void _draw_nema_gfx_tile(lv_draw_task_t * t, const lv_draw_image_dsc_t * dsc, const lv_area_t * coords)
{
lv_image_decoder_dsc_t decoder_dsc;
lv_result_t res = lv_image_decoder_open(&decoder_dsc, dsc->src, NULL);
if(res != LV_RESULT_OK) {
LV_LOG_ERROR("Failed to open image");
return;
}
int32_t img_w = dsc->header.w;
int32_t img_h = dsc->header.h;
lv_area_t tile_area;
if(lv_area_get_width(&dsc->image_area) >= 0) {
tile_area = dsc->image_area;
}
else {
tile_area = *coords;
}
lv_area_set_width(&tile_area, img_w);
lv_area_set_height(&tile_area, img_h);
int32_t tile_x_start = tile_area.x1;
while(tile_area.y1 <= t->clip_area.y2) {
while(tile_area.x1 <= t->clip_area.x2) {
lv_area_t clipped_img_area;
if(lv_area_intersect(&clipped_img_area, &tile_area, &t->clip_area)) {
_draw_nema_gfx_img(t, dsc, &tile_area);
}
tile_area.x1 += img_w;
tile_area.x2 += img_w;
}
tile_area.y1 += img_h;
tile_area.y2 += img_h;
tile_area.x1 = tile_x_start;
tile_area.x2 = tile_x_start + img_w - 1;
}
lv_image_decoder_close(&decoder_dsc);
}
static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * dsc, const lv_area_t * coords)
static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * dsc,
const lv_image_decoder_dsc_t * decoder_dsc, lv_draw_image_sup_t * sup,
const lv_area_t * img_coords, const lv_area_t * clipped_img_area)
{
if(dsc->opa <= LV_OPA_MIN) return;
lv_draw_nema_gfx_unit_t * draw_nema_gfx_unit = (lv_draw_nema_gfx_unit_t *)t->draw_unit;
lv_layer_t * layer = t->target_layer;
const lv_image_dsc_t * img_dsc = dsc->src;
const lv_draw_buf_t * decoded = decoder_dsc->decoded;
const lv_image_header_t * header = &decoded->header;
bool masked = dsc->bitmap_mask_src != NULL;
lv_area_t blend_area;
/*Let's get the blend area which is the intersection of the area to fill and the clip area.*/
if(!lv_area_intersect(&blend_area, coords, &t->clip_area))
return; /*Fully clipped, nothing to do*/
lv_area_t rel_clip_area;
lv_area_copy(&rel_clip_area, &t->clip_area);
lv_area_copy(&rel_clip_area, clipped_img_area);
lv_area_move(&rel_clip_area, -layer->buf_area.x1, -layer->buf_area.y1);
bool has_transform = (dsc->rotation != 0 || dsc->scale_x != LV_SCALE_NONE || dsc->scale_y != LV_SCALE_NONE);
bool recolor = (dsc->recolor_opa > LV_OPA_MIN);
/*Make the blend area relative to the buffer*/
lv_area_move(&blend_area, -layer->buf_area.x1, -layer->buf_area.y1);
uint32_t tex_w = lv_area_get_width(coords);
uint32_t tex_h = lv_area_get_height(coords);
uint32_t tex_w = header->w;
uint32_t tex_h = header->h;
nema_set_clip(rel_clip_area.x1, rel_clip_area.y1, lv_area_get_width(&rel_clip_area),
lv_area_get_height(&rel_clip_area));
@@ -132,11 +81,11 @@ static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * d
lv_color_format_t dst_cf = layer->draw_buf->header.cf;
uint32_t dst_nema_cf = lv_nemagfx_cf_to_nema(dst_cf);
const void * src_buf = img_dsc->data;
const void * src_buf = decoded->data;
uint32_t blending_mode = lv_nemagfx_blending_mode(dsc->blend_mode);
lv_color_format_t src_cf = img_dsc->header.cf;
lv_color_format_t src_cf = header->cf;
/*Image contains Alpha*/
if(src_cf == LV_COLOR_FORMAT_ARGB8888 || src_cf == LV_COLOR_FORMAT_XRGB8888) {
@@ -146,7 +95,7 @@ static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * d
uint32_t src_nema_cf = lv_nemagfx_cf_to_nema(src_cf);
/* the stride should be computed internally for NEMA_TSC images and images missing a stride value */
int32_t src_stride = (src_cf >= LV_COLOR_FORMAT_NEMA_TSC_START && src_cf <= LV_COLOR_FORMAT_NEMA_TSC_END)
|| img_dsc->header.stride == 0 ? -1 : (int32_t)img_dsc->header.stride;
|| header->stride == 0 ? -1 : (int32_t)header->stride;
int32_t stride = (dst_cf >= LV_COLOR_FORMAT_NEMA_TSC_START && dst_cf <= LV_COLOR_FORMAT_NEMA_TSC_END) ?
-1 : lv_area_get_width(&(layer->buf_area)) * lv_color_format_get_size(dst_cf);
@@ -189,13 +138,13 @@ static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * d
const lv_area_t * image_area;
lv_area_t mask_area;
if(lv_area_get_width(&dsc->image_area) < 0) image_area = coords;
if(lv_area_get_width(&dsc->image_area) < 0) image_area = img_coords;
else image_area = &dsc->image_area;
lv_area_set(&mask_area, 0, 0, dsc->bitmap_mask_src->header.w - 1, dsc->bitmap_mask_src->header.h - 1);
lv_area_align(image_area, &mask_area, LV_ALIGN_CENTER, 0, 0);
mask_buf += dsc->bitmap_mask_src->header.w * (coords->y1 - mask_area.y1) + (coords->x1 - mask_area.x1);
mask_buf += dsc->bitmap_mask_src->header.w * (img_coords->y1 - mask_area.y1) + (img_coords->x1 - mask_area.x1);
nema_bind_tex(NEMA_TEX3, (uintptr_t)NEMA_VIRT2PHYS(mask_buf), mask->header.w, mask->header.h,
lv_nemagfx_mask_cf_to_nema(mask->header.cf),
@@ -206,13 +155,13 @@ static void _draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * d
nema_set_blend_blit(blending_mode);
if(!has_transform) {
nema_blit_rect((coords->x1 - layer->buf_area.x1),
(coords->y1 - layer->buf_area.y1), tex_w, tex_h);
nema_blit_rect((img_coords->x1 - layer->buf_area.x1),
(img_coords->y1 - layer->buf_area.y1), tex_w, tex_h);
}
else {
/*Calculate the transformed points*/
float x0 = (coords->x1 - layer->buf_area.x1);
float y0 = (coords->y1 - layer->buf_area.y1);
float x0 = (img_coords->x1 - layer->buf_area.x1);
float y0 = (img_coords->y1 - layer->buf_area.y1);
float x1 = x0 + tex_w ;
float y1 = y0;
float x2 = x0 + tex_w ;
@@ -273,10 +222,10 @@ void lv_draw_nema_gfx_img(lv_draw_task_t * t, const lv_draw_image_dsc_t * dsc, c
{
if(!dsc->tile) {
_draw_nema_gfx_img(t, dsc, coords);
lv_draw_image_normal_helper(t, dsc, coords, _draw_nema_gfx_img);
}
else {
_draw_nema_gfx_tile(t, dsc, coords);
lv_draw_image_tiled_helper(t, dsc, coords, _draw_nema_gfx_img);
}
}