Update for splatw0q->splatw3q change

3 files changed, 120 insertions, 31 deletions

diff --git a/pixman/pixman-orc.c b/pixman/pixman-orc.c
index 6a1da67..df00568 100644
--- a/pixman/pixman-orc.c
+++ b/pixman/pixman-orc.c
@@ -268,8 +268,10 @@ orc_composite_add_8888_8888 (pixman_implementation_t *imp,
                                  int32_t                  width,
                                  int32_t                  height)
 {
-    uint32_t    *src, *src_line;
-    uint32_t    *dst, *dst_line;
+    //uint32_t    *src, *src_line;
+    //uint32_t    *dst, *dst_line;
+    uint32_t    *src_line;
+    uint32_t    *dst_line;
     int src_stride, dst_stride;
 
     PIXMAN_IMAGE_GET_LINE (
@@ -277,6 +279,9 @@ orc_composite_add_8888_8888 (pixman_implementation_t *imp,
     PIXMAN_IMAGE_GET_LINE (
 	src_image, src_x, src_y, uint32_t, src_stride, src_line, 1);
 
+    orc_composite_add_8888_8888_2d (dst_line, dst_stride * sizeof(uint32_t),
+        src_line, src_stride * sizeof(uint32_t), width, height);
+#if 0
     while (height--) {
         src = src_line;
         src_line += src_stride;
@@ -285,6 +290,7 @@ orc_composite_add_8888_8888 (pixman_implementation_t *imp,
 
         orc_composite_add_8888_8888_line (dst, src, width);
     }
+#endif
 }
 
 
@@ -398,6 +404,39 @@ orc_composite_over_8888_8_8888 (pixman_implementation_t *imp,
     }
 }
 
+static void
+orc_composite_over_n_8888_8888_ca (pixman_implementation_t *imp,
+                                 pixman_op_t              op,
+                                 pixman_image_t *         src_image,
+                                 pixman_image_t *         mask_image,
+                                 pixman_image_t *         dst_image,
+                                 int32_t                  src_x,
+                                 int32_t                  src_y,
+                                 int32_t                  mask_x,
+                                 int32_t                  mask_y,
+                                 int32_t                  dest_x,
+                                 int32_t                  dest_y,
+                                 int32_t                  width,
+                                 int32_t                  height)
+{
+    uint32_t    *dst_line;
+    uint32_t    *mask_line;
+    int dst_stride, mask_stride;
+    uint32_t src;
+
+    src = _pixman_image_get_solid (src_image, dst_image->bits.format);
+
+    if (src == 0) return;
+
+    PIXMAN_IMAGE_GET_LINE (
+	dst_image, dest_x, dest_y, uint32_t, dst_stride, dst_line, 1);
+    PIXMAN_IMAGE_GET_LINE (
+	mask_image, mask_x, mask_y, uint32_t, mask_stride, mask_line, 1);
+
+    orc_composite_over_n_8888_8888_ca_2d (dst_line,
+        dst_stride * sizeof(uint32_t), mask_line,
+        mask_stride * sizeof(uint32_t), src, width, height);
+}
 
 
 static const pixman_fast_path_t orc_fast_paths[] =
@@ -437,10 +476,12 @@ static const pixman_fast_path_t orc_fast_paths[] =
     PIXMAN_STD_FAST_PATH( OVER, a8r8g8b8, solid,    x8r8g8b8, orc_composite_over_8888_n_8888    ),
     PIXMAN_STD_FAST_PATH( OVER, a8b8g8r8, solid,    a8b8g8r8, orc_composite_over_8888_n_8888    ),
     PIXMAN_STD_FAST_PATH( OVER, a8b8g8r8, solid,    x8b8g8r8, orc_composite_over_8888_n_8888    ),
-    PIXMAN_STD_FAST_PATH( OVER, solid, a8r8g8b8_ca, a8r8g8b8, orc_composite_over_n_8888_8888_ca ),
-    PIXMAN_STD_FAST_PATH( OVER, solid, a8r8g8b8_ca, x8r8g8b8, orc_composite_over_n_8888_8888_ca ),
-    PIXMAN_STD_FAST_PATH( OVER, solid, a8b8g8r8_ca, a8b8g8r8, orc_composite_over_n_8888_8888_ca ),
-    PIXMAN_STD_FAST_PATH( OVER, solid, a8b8g8r8_ca, x8b8g8r8, orc_composite_over_n_8888_8888_ca ),
+#endif
+    PIXMAN_STD_FAST_PATH_CA( OVER, solid, a8r8g8b8, a8r8g8b8, orc_composite_over_n_8888_8888_ca ),
+    PIXMAN_STD_FAST_PATH_CA( OVER, solid, a8r8g8b8, x8r8g8b8, orc_composite_over_n_8888_8888_ca ),
+    PIXMAN_STD_FAST_PATH_CA( OVER, solid, a8b8g8r8, a8b8g8r8, orc_composite_over_n_8888_8888_ca ),
+    PIXMAN_STD_FAST_PATH_CA( OVER, solid, a8b8g8r8, x8b8g8r8, orc_composite_over_n_8888_8888_ca ),
+#if 0
     PIXMAN_STD_FAST_PATH( OVER, solid, a8r8g8b8_ca, r5g6b5,   orc_composite_over_n_8888_0565_ca ),
     PIXMAN_STD_FAST_PATH( OVER, solid, a8b8g8r8_ca, b5g6r5,   orc_composite_over_n_8888_0565_ca ),
     PIXMAN_STD_FAST_PATH( OVER, pixbuf,   pixbuf,   a8r8g8b8, orc_composite_over_pixbuf_8888    ),
diff --git a/pixman/pixman-orccode.orc b/pixman/pixman-orccode.orc
index d39c9a8..beec52c 100644
--- a/pixman/pixman-orccode.orc
+++ b/pixman/pixman-orccode.orc
@@ -46,7 +46,16 @@ copyw d1, s1
 copyl d1, s1
 
 
-.function orc_composite_add_8888_8888_line
+.function orc_copy_u8_2d
+.dest 1 d1
+.source 1 s1
+.flags 2d
+
+copyb d1, s1
+
+
+.function orc_composite_add_8888_8888_2d
+.flags 2d
 .dest 4 d1
 .source 4 s1
 
@@ -85,7 +94,7 @@ addusb d1, d1, t2
 
 x4 convubw t1, s1
 x4 convubw t2, s2
-splatw0q t2, t2
+splatw3q t2, t2
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 x4 convwb t3, t1
@@ -113,7 +122,7 @@ x4 addusb d1, d1, s1
 #compover d1, d1, t1
 x4 convubw t1, s1
 x4 convubw t2, s2
-splatw0q t2, t2
+splatw3q t2, t2
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 x4 convwb t3, t1
@@ -121,7 +130,7 @@ x4 convwb t3, t1
 loadl d, d1
 x4 convubw d_wide, d
 x4 xorw t1, t1, 0x00ff
-splatw0q t2, t1
+splatw3q t2, t1
 x4 mullw t1, d_wide, t2
 x4 div255w t1, t1
 x4 convwb d, t1
@@ -142,7 +151,7 @@ x4 convubw t1, s
 loadl d, d1
 x4 convubw d_wide, d
 x4 xorw t1, t1, 0x00ff
-splatw0q t2, t1
+splatw3q t2, t1
 x4 mullw t1, d_wide, t2
 x4 div255w t1, t1
 x4 convwb d, t1
@@ -161,12 +170,12 @@ x4 addusb d1, d, s
 
 x4 convubw t1, s1
 x4 convubw t2, s2
-splatw0q t2, t2
+splatw3q t2, t2
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 # ORC_MULDIV_255((s),(m)), m is from dest
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 x4 convwb d1, t1
@@ -185,7 +194,7 @@ x4 convwb d1, t1
 x4 convubw t1, s1
 # ORC_MULDIV_255((s),(m)), m is from dest
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 x4 convwb d1, t1
@@ -203,12 +212,12 @@ x4 convwb d1, t1
 
 x4 convubw t1, s1
 x4 convubw t2, s2
-splatw0q t2, t2
+splatw3q t2, t2
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 # ORC_MULDIV_255((s),(m)), m is from dest
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 xorw t2, t2, 0x00ff
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
@@ -228,7 +237,7 @@ x4 convwb d1, t1
 x4 convubw t1, s1
 # ORC_MULDIV_255((s),(m)), m is from dest
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 xorw t2, t2, 0x00ff
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
@@ -252,18 +261,18 @@ x4 convwb d1, t1
 
 x4 convubw t1, s1
 x4 convubw t2, s2
-splatw0q t2, t2
+splatw3q t2, t2
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 mullw t3, t1, t2
 x4 div255w t3, t3
 x4 convwb t4, t3
 
 x4 convubw d_wide, d1
-splatw0q t2, t1
+splatw3q t2, t1
 x4 xorw t2, t2, 0x00ff
 x4 mullw t1, d_wide, t2
 x4 div255w t1, t1
@@ -287,13 +296,13 @@ x4 addusb d1, t4, t5
 x4 convubw t1, s1
 
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 mullw t3, t1, t2
 x4 div255w t3, t3
 x4 convwb t4, t3
 
 x4 convubw d_wide, d1
-splatw0q t2, t1
+splatw3q t2, t1
 x4 xorw t2, t2, 0x00ff
 x4 mullw t1, d_wide, t2
 x4 div255w t1, t1
@@ -318,19 +327,19 @@ x4 addusb d1, t4, t5
 
 x4 convubw t1, s1
 x4 convubw t2, s2
-splatw0q t2, t2
+splatw3q t2, t2
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 xorw t2, t2, 0x00ff
 x4 mullw t3, t1, t2
 x4 div255w t3, t3
 x4 convwb t4, t3
 
 x4 convubw d_wide, d1
-splatw0q t2, t1
+splatw3q t2, t1
 x4 xorw t2, t2, 0x00ff
 x4 mullw t1, d_wide, t2
 x4 div255w t1, t1
@@ -353,14 +362,14 @@ x4 addusb d1, t4, t5
 
 x4 convubw t1, s1
 x4 convubw d_wide, d1
-splatw0q t2, d_wide
+splatw3q t2, d_wide
 x4 xorw t2, t2, 0x00ff
 x4 mullw t3, t1, t2
 x4 div255w t3, t3
 x4 convwb t4, t3
 
 x4 convubw d_wide, d1
-splatw0q t2, t1
+splatw3q t2, t1
 x4 xorw t2, t2, 0x00ff
 x4 mullw t1, d_wide, t2
 x4 div255w t1, t1
@@ -380,7 +389,7 @@ x4 addusb d1, t4, t5
 
 x4 convubw t1, s1
 x4 convubw t2, s2
-#splatw0q t2, t2
+#splatw3q t2, t2
 x4 mullw t1, t1, t2
 x4 div255w t1, t1
 x4 convwb t3, t1
@@ -409,7 +418,7 @@ x4 addusb d1, d1, s1
 
 x4 convubw s_wide, s1
 x4 convubw m_wide, s2
-splatw0q xa, s_wide
+splatw3q xa, s_wide
 x4 mullw s_wide, s_wide, m_wide
 x4 div255w s_wide, s_wide
 x4 convwb s, s_wide
@@ -438,7 +447,7 @@ x4 addusb d1, d, s
 .temp 4 s
 
 x4 convubw s_wide, s1
-splatw0q xa, s_wide
+splatw3q xa, s_wide
 x4 convwb s, s_wide
 x4 copyw m_wide, xa
 loadl d, d1
@@ -471,9 +480,41 @@ x4 convwb t3, t1
 loadl d, d1
 x4 convubw d_wide, d
 x4 xorw t1, t1, 0x00ff
-splatw0q t2, t1
+splatw3q t2, t1
 x4 mullw t1, d_wide, t2
 x4 div255w t1, t1
 x4 convwb d, t1
 x4 addusb d1, d, t3
 
+
+
+.function orc_composite_over_n_8888_8888_ca_2d
+.flags 2d
+.dest 4 d1
+.source 4 s1
+.param 4 p1
+.temp 8 t1
+.temp 8 t2
+.temp 4 d
+.temp 8 d_wide
+.temp 8 m_wide
+.temp 8 s_wide
+.temp 8 xa
+.temp 4 s
+
+x4 convubw s_wide, p1
+x4 convubw m_wide, s1
+splatw3q xa, s_wide
+x4 mullw s_wide, s_wide, m_wide
+x4 div255w s_wide, s_wide
+x4 convwb s, s_wide
+x4 mullw m_wide, m_wide, xa
+x4 div255w m_wide, m_wide
+loadl d, d1
+x4 convubw d_wide, d
+x4 xorw m_wide, m_wide, 0x00ff
+x4 mullw t1, d_wide, m_wide
+x4 div255w t1, t1
+x4 convwb d, t1
+x4 addusb d1, d, s
+
diff --git a/test/Makefile.am b/test/Makefile.am
index 80f78d7..031b70e 100644
--- a/test/Makefile.am
+++ b/test/Makefile.am
@@ -22,6 +22,7 @@ TESTPROGRAMS =			\
 	scaling-test		\
 	affine-test		\
 	composite		\
+	pixmanorc		\
 	ds
 
 pdf_op_test_SOURCES = pdf-op-test.c utils.c utils.h
@@ -41,6 +42,12 @@ gradient_crash_test_SOURCES = gradient-crash-test.c utils.c utils.h
 stress_test_SOURCES = stress-test.c utils.c utils.h
 scaling_helpers_test_SOURCES = scaling-helpers-test.c utils.c utils.h
 
+pixmanorc_LDADD = $(TEST_LDADD) -lorc-test-0.4
+pixmanorc_SOURCES = pixmanorc.c
+
+pixmanorc.c: $(top_srcdir)/pixman/pixman-orccode.orc
+	$(ORCC) --test -o $@ $<
+
 # Benchmarks
 
 BENCHMARKS =			\