1 files changed, 447 insertions, 0 deletions

diff --git a/libraries/evas/src/lib/engines/common_8/evas_soft8_scanline_blend.c b/libraries/evas/src/lib/engines/common_8/evas_soft8_scanline_blend.c
new file mode 100644
index 0000000..5fc62c2
--- /dev/null
+++ b/libraries/evas/src/lib/engines/common_8/evas_soft8_scanline_blend.c
@@ -0,0 +1,447 @@
+/** NOTE: This file is meant to be included by users **/
+
+/** NOTE2: r, g, b parameters are 16bits, so you can pass 0 to 256 inclusive.
+ **        this is due our division by 256 when multiplying the color.
+ **/
+
+/*****************************************************************************
+ * Scanline processing
+ *
+ *    _soft8_scanline_<description>_<src>_<dst>[_<modifier>]()
+ *
+ ****************************************************************************/
+
+EFL_ALWAYS_INLINE void
+_soft8_pt_blend_transp_solid(DATA8 * p_dst, DATA8 src, DATA8 alpha)
+{
+   if (alpha == 0xff)
+      *p_dst = src;
+   else if (alpha != 0)
+     {
+        *p_dst = GRY_8_BLEND(src, *p_dst, alpha);
+     }
+}
+
+/***********************************************************************
+ * Regular blend operations
+ */
+static void
+_soft8_scanline_blend_transp_solid(DATA8 * src, DATA8 * alpha, DATA8 * dst,
+                                   int size)
+{
+   DATA8 *start, *end;
+
+   start = dst;
+   end = start + (size & ~7);
+
+   pld(alpha, 0);
+   pld(src, 0);
+
+   /* work on 8 pixels per time, do data preload */
+   while (start < end)
+     {
+        DATA8 alpha1, alpha2;
+
+        alpha1 = alpha[0];
+        alpha += 8;
+
+        /* empirical tests show these give the best performance */
+        pld(alpha, 8);
+        pld(src, 32);
+
+        src += 8;
+        start += 8;
+
+        alpha2 = alpha[-7];
+        _soft8_pt_blend_transp_solid(start - 8, src[-8], alpha1);
+
+        alpha1 = alpha[-6];
+        _soft8_pt_blend_transp_solid(start - 7, src[-7], alpha2);
+
+        alpha2 = alpha[-5];
+        _soft8_pt_blend_transp_solid(start - 6, src[-6], alpha1);
+
+        alpha1 = alpha[-4];
+        _soft8_pt_blend_transp_solid(start - 5, src[-5], alpha2);
+
+        alpha2 = alpha[-3];
+        _soft8_pt_blend_transp_solid(start - 4, src[-4], alpha1);
+
+        alpha1 = alpha[-2];
+        _soft8_pt_blend_transp_solid(start - 3, src[-3], alpha2);
+
+        alpha2 = alpha[-1];
+        _soft8_pt_blend_transp_solid(start - 2, src[-2], alpha1);
+
+        _soft8_pt_blend_transp_solid(start - 1, src[-1], alpha2);
+     }
+
+   /* remaining pixels (up to 7) */
+   end = start + (size & 7);
+   for (; start < end; start++, src++, alpha++)
+      _soft8_pt_blend_transp_solid(start, *src, *alpha);
+}
+
+EFL_ALWAYS_INLINE void
+_soft8_pt_blend_solid_solid(DATA8 * p_dst, DATA8 src)
+{
+   *p_dst = src;
+}
+
+static inline void
+_soft8_scanline_blend_solid_solid(DATA8 * src, DATA8 * dst, int size)
+{
+   memcpy(dst, src, size * sizeof(DATA8));
+}
+
+/***********************************************************************
+ * Blend operations taking an extra alpha (fade in, out)
+ */
+
+static inline void
+_soft8_pt_blend_transp_solid_mul_alpha(DATA8 * p_dst, DATA8 src, DATA8 alpha,
+                                       DATA8 rel_alpha)
+{
+   alpha = alpha * rel_alpha;
+   if (alpha == 0)
+      return;
+
+   alpha++;
+   *p_dst = GRY_8_BLEND((src * rel_alpha) & 0xff, *p_dst, alpha);
+}
+
+static void
+_soft8_scanline_blend_transp_solid_mul_alpha(DATA8 * src, DATA8 * alpha,
+                                             DATA8 * dst, int size,
+                                             const DATA8 rel_alpha)
+{
+   DATA8 *start, *end;
+
+   start = dst;
+   end = start + (size & ~7);
+
+   pld(alpha, 0);
+   pld(src, 0);
+
+   while (start < end)
+     {
+        DATA8 alpha1, alpha2;
+
+        alpha1 = alpha[0];
+        alpha += 8;
+
+        pld(alpha, 8);
+        pld(src, 32);
+
+        src += 8;
+        start += 8;
+
+        alpha2 = alpha[-7];
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 8, src[-8], alpha1, rel_alpha);
+
+        alpha1 = alpha[-6];
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 7, src[-7], alpha2, rel_alpha);
+
+        alpha2 = alpha[-5];
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 6, src[-6], alpha1, rel_alpha);
+
+        alpha1 = alpha[-4];
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 5, src[-5], alpha2, rel_alpha);
+
+        alpha2 = alpha[-3];
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 4, src[-4], alpha1, rel_alpha);
+
+        alpha1 = alpha[-2];
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 3, src[-3], alpha2, rel_alpha);
+
+        alpha2 = alpha[-1];
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 2, src[-2], alpha1, rel_alpha);
+
+        _soft8_pt_blend_transp_solid_mul_alpha
+            (start - 1, src[-1], alpha2, rel_alpha);
+     }
+
+   end = start + (size & 7);
+   for (; start < end; start++, src++, alpha++)
+      _soft8_pt_blend_transp_solid_mul_alpha(start, *src, *alpha, rel_alpha);
+}
+
+EFL_ALWAYS_INLINE void
+_soft8_pt_blend_solid_solid_mul_alpha(DATA8 * p_dst, DATA8 src, DATA8 rel_alpha)
+{
+   *p_dst = GRY_8_BLEND_UNMUL(src, *p_dst, rel_alpha);
+}
+
+static void
+_soft8_scanline_blend_solid_solid_mul_alpha(DATA8 * src, DATA8 * dst, int size,
+                                            DATA8 rel_alpha)
+{
+   DATA8 *start, *end;
+
+   start = dst;
+   end = start + (size & ~7);
+
+   pld(src, 0);
+
+   while (start < end)
+     {
+        pld(src, 32);
+        UNROLL8(
+                  {
+                  _soft8_pt_blend_solid_solid_mul_alpha(start, *src, rel_alpha);
+                  start++; src++;}
+        );
+     }
+
+   end = start + (size & 7);
+   for (; start < end; start++, src++)
+      _soft8_pt_blend_solid_solid_mul_alpha(start, *src, rel_alpha);
+}
+
+/***********************************************************************
+ * Blend operations with extra alpha and multiply color
+ */
+
+EFL_ALWAYS_INLINE void
+_soft8_pt_blend_transp_solid_mul_color_transp(DATA8 * p_dst, DATA8 src,
+                                              DATA8 alpha, DATA8 rel_alpha,
+                                              DATA8 r, DATA8 g, DATA8 b)
+{
+   alpha = alpha * rel_alpha;
+   if (alpha == 0)
+      return;
+
+   alpha++;
+
+   DATA8 gry8 = (src * GRY_8_FROM_COMPONENTS(r, g, b)) >> 8;
+   *p_dst = GRY_8_BLEND(gry8, *p_dst, alpha);
+}
+
+static void
+_soft8_scanline_blend_transp_solid_mul_color_transp(DATA8 * src, DATA8 * alpha,
+                                                    DATA8 * dst, int size,
+                                                    DATA8 rel_alpha, DATA8 r,
+                                                    DATA8 g, DATA8 b)
+{
+   DATA8 *start, *end;
+
+   start = dst;
+   end = start + (size & ~7);
+
+   pld(alpha, 0);
+   pld(src, 0);
+
+   while (start < end)
+     {
+        DATA8 alpha1, alpha2;
+
+        alpha1 = alpha[0];
+        alpha += 8;
+
+        pld(src, 32);
+        pld(start, 32);
+
+        src += 8;
+        start += 8;
+
+        alpha2 = alpha[-7];
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 8, src[-8], alpha1, rel_alpha, r, g, b);
+
+        alpha1 = alpha[-6];
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 7, src[-7], alpha2, rel_alpha, r, g, b);
+
+        alpha2 = alpha[-5];
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 6, src[-6], alpha1, rel_alpha, r, g, b);
+
+        alpha1 = alpha[-4];
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 5, src[-5], alpha2, rel_alpha, r, g, b);
+
+        alpha2 = alpha[-3];
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 4, src[-4], alpha1, rel_alpha, r, g, b);
+
+        alpha1 = alpha[-2];
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 3, src[-3], alpha2, rel_alpha, r, g, b);
+
+        alpha2 = alpha[-1];
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 2, src[-2], alpha1, rel_alpha, r, g, b);
+
+        _soft8_pt_blend_transp_solid_mul_color_transp
+            (start - 1, src[-1], alpha2, rel_alpha, r, g, b);
+     }
+
+   end = start + (size & 7);
+   for (; start < end; start++, src++, alpha++)
+      _soft8_pt_blend_transp_solid_mul_color_transp
+          (start, *src, *alpha, rel_alpha, r, g, b);
+}
+
+EFL_ALWAYS_INLINE void
+_soft8_pt_blend_solid_solid_mul_color_transp(DATA8 * p_dst, DATA8 src,
+                                             DATA8 rel_alpha, DATA8 r, DATA8 g,
+                                             DATA8 b)
+{
+   DATA8 gry8 = (src * GRY_8_FROM_COMPONENTS(r, g, b)) >> 8;
+   *p_dst = GRY_8_BLEND(gry8, *p_dst, rel_alpha);
+}
+
+static void
+_soft8_scanline_blend_solid_solid_mul_color_transp(DATA8 * src, DATA8 * dst,
+                                                   int size, DATA8 rel_alpha,
+                                                   DATA8 r, DATA8 g, DATA8 b)
+{
+   DATA8 *start, *end;
+
+   start = dst;
+   end = start + (size & ~7);
+
+   pld(src, 0);
+
+   while (start < end)
+     {
+        pld(src, 32);
+        UNROLL8(
+                  {
+                  _soft8_pt_blend_solid_solid_mul_color_transp
+                  (start, *src, rel_alpha, r, g, b); start++; src++;}
+        );
+     }
+
+   end = start + (size & 7);
+   for (; start < end; start++, src++)
+      _soft8_pt_blend_solid_solid_mul_color_transp
+          (start, *src, rel_alpha, r, g, b);
+}
+
+/***********************************************************************
+ * Blend operations with extra multiply color
+ */
+EFL_ALWAYS_INLINE void
+_soft8_pt_blend_transp_solid_mul_color_solid(DATA8 * p_dst, DATA8 src,
+                                             DATA8 alpha, DATA8 r, DATA8 g,
+                                             DATA8 b)
+{
+   if (alpha == 0)
+      return;
+
+   DATA8 gry8 = (src * GRY_8_FROM_COMPONENTS(r, g, b)) >> 8;
+
+   if (alpha == 0xff)
+      *p_dst = gry8;
+   else
+     {
+        *p_dst = GRY_8_BLEND(gry8, *p_dst, alpha);
+     }
+}
+
+static void
+_soft8_scanline_blend_transp_solid_mul_color_solid(DATA8 * src, DATA8 * alpha,
+                                                   DATA8 * dst, int size,
+                                                   DATA8 r, DATA8 g, DATA8 b)
+{
+   DATA8 *start, *end;
+
+   start = dst;
+   end = start + (size & ~7);
+
+   pld(alpha, 0);
+   pld(src, 0);
+
+   while (start < end)
+     {
+        DATA8 alpha1, alpha2;
+
+        alpha1 = alpha[0];
+        alpha += 8;
+
+        pld(alpha, 8);
+        pld(src, 32);
+
+        src += 8;
+        start += 8;
+
+        alpha2 = alpha[-7];
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 8, src[-8], alpha1, r, g, b);
+
+        alpha1 = alpha[-6];
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 7, src[-7], alpha2, r, g, b);
+
+        alpha2 = alpha[-5];
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 6, src[-6], alpha1, r, g, b);
+
+        alpha1 = alpha[-4];
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 5, src[-5], alpha2, r, g, b);
+
+        alpha2 = alpha[-3];
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 4, src[-4], alpha1, r, g, b);
+
+        alpha1 = alpha[-2];
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 3, src[-3], alpha2, r, g, b);
+
+        alpha2 = alpha[-1];
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 2, src[-2], alpha1, r, g, b);
+
+        _soft8_pt_blend_transp_solid_mul_color_solid
+            (start - 1, src[-1], alpha2, r, g, b);
+     }
+
+   end = start + (size & 7);
+   for (; start < end; start++, src++, alpha++)
+      _soft8_pt_blend_transp_solid_mul_color_solid
+          (start, *src, *alpha, r, g, b);
+}
+
+EFL_ALWAYS_INLINE void
+_soft8_pt_blend_solid_solid_mul_color_solid(DATA8 * p_dst, DATA8 src, DATA8 r,
+                                            DATA8 g, DATA8 b)
+{
+   *p_dst = (src * GRY_8_FROM_COMPONENTS(r, g, b)) >> 8;
+}
+
+static void
+_soft8_scanline_blend_solid_solid_mul_color_solid(DATA8 * src, DATA8 * dst,
+                                                  int size, DATA8 r, DATA8 g,
+                                                  DATA8 b)
+{
+   DATA8 *start, *end;
+
+   start = dst;
+   end = start + (size & ~7);
+
+   pld(src, 0);
+
+   while (start < end)
+     {
+        pld(src, 32);
+        UNROLL8(
+                  {
+                  _soft8_pt_blend_solid_solid_mul_color_solid(start, *src, r, g,
+                                                              b); start++;
+                  src++;}
+        );
+     }
+
+   end = start + (size & 7);
+   for (; start < end; start++, src++)
+      _soft8_pt_blend_solid_solid_mul_color_solid(start, *src, r, g, b);
+}