[Pixman] [PATCH 6/5] mmx: add over_reverse_n_8888

[Matt Turner]

Loongson:
over_reverse_n_8888 =  L1:  16.04  L2:  15.35  M: 10.20 ( 27.96%)  HT: 10.95  VT: 10.45  R:  9.18  RT:  6.99 (  76Kops/s)
over_reverse_n_8888 =  L1:  27.40  L2:  26.67  M: 16.97 ( 45.78%)  HT: 16.66  VT: 15.38  R: 14.15  RT:  9.44 (  97Kops/s)

image                      poppler   34.106   35.500   1.48%    6/6
image                      poppler   29.598   30.835   1.70%    6/6

ARM/iwMMXt:
over_reverse_n_8888 =  L1:  15.63  L2:  14.33  M: 10.83 ( 27.55%)  HT:  9.78  VT:  9.91  R:  9.49  RT:  6.96 (  69Kops/s)
over_reverse_n_8888 =  L1:  22.79  L2:  19.40  M: 13.76 ( 34.19%)  HT: 11.66  VT: 11.86  R: 11.17  RT:  7.85 (  75Kops/s)

image                      poppler   38.040   38.606   1.10%    6/6
image                      poppler   31.686   32.278   0.80%    5/6
---
 pixman/pixman-mmx.c       |   72 +++++++++++++++++++++++++++++++++++++++++++++
 test/lowlevel-blt-bench.c |    1 +
 2 files changed, 73 insertions(+), 0 deletions(-)

diff --git a/pixman/pixman-mmx.c b/pixman/pixman-mmx.c
index 80cdf36..6413ded 100644
--- a/pixman/pixman-mmx.c
+++ b/pixman/pixman-mmx.c
@@ -3430,6 +3430,75 @@ mmx_composite_over_x888_8_8888 (pixman_implementation_t *imp,
     _mm_empty ();
 }
 
+static void
+mmx_composite_over_reverse_n_8888 (pixman_implementation_t *imp,
+                                   pixman_composite_info_t *info)
+{
+    PIXMAN_COMPOSITE_ARGS (info);
+    uint32_t src;
+    uint32_t    *dst_line, *dst;
+    int32_t w;
+    int dst_stride;
+    __m64 vsrc;
+
+    CHECKPOINT ();
+
+    src = _pixman_image_get_solid (imp, src_image, dest_image->bits.format);
+
+    if (src == 0)
+	return;
+
+    PIXMAN_IMAGE_GET_LINE (dest_image, dest_x, dest_y, uint32_t, dst_stride, dst_line, 1);
+
+    vsrc = load8888 (&src);
+
+    while (height--)
+    {
+	dst = dst_line;
+	dst_line += dst_stride;
+	w = width;
+
+	CHECKPOINT ();
+
+	while (w && (unsigned long)dst & 7)
+	{
+	    __m64 vdest = load8888 (dst);
+
+	    store8888 (dst, over (vdest, expand_alpha (vdest), vsrc));
+
+	    w--;
+	    dst++;
+	}
+
+	while (w >= 2)
+	{
+	    __m64 vdest = *(__m64 *)dst;
+	    __m64 dest0 = expand8888 (vdest, 0);
+	    __m64 dest1 = expand8888 (vdest, 1);
+
+
+	    dest0 = over (dest0, expand_alpha (dest0), vsrc);
+	    dest1 = over (dest1, expand_alpha (dest1), vsrc);
+
+	    *(__m64 *)dst = pack8888 (dest0, dest1);
+
+	    dst += 2;
+	    w -= 2;
+	}
+
+	CHECKPOINT ();
+
+	if (w)
+	{
+	    __m64 vdest = load8888 (dst);
+
+	    store8888 (dst, over (vdest, expand_alpha (vdest), vsrc));
+	}
+    }
+
+    _mm_empty ();
+}
+
 static uint32_t *
 mmx_fetch_x8r8g8b8 (pixman_iter_t *iter, const uint32_t *mask)
 {
@@ -3658,6 +3727,9 @@ static const pixman_fast_path_t mmx_fast_paths[] =
     PIXMAN_STD_FAST_PATH    (OVER, a8b8g8r8, null,     x8b8g8r8, mmx_composite_over_8888_8888      ),
     PIXMAN_STD_FAST_PATH    (OVER, a8b8g8r8, null,     b5g6r5,   mmx_composite_over_8888_0565      ),
 
+    PIXMAN_STD_FAST_PATH    (OVER_REVERSE, solid, null, a8r8g8b8, mmx_composite_over_reverse_n_8888),
+    PIXMAN_STD_FAST_PATH    (OVER_REVERSE, solid, null, a8b8g8r8, mmx_composite_over_reverse_n_8888),
+
     PIXMAN_STD_FAST_PATH    (ADD,  r5g6b5,   null,     r5g6b5,   mmx_composite_add_0565_0565       ),
     PIXMAN_STD_FAST_PATH    (ADD,  b5g6r5,   null,     b5g6r5,   mmx_composite_add_0565_0565       ),
     PIXMAN_STD_FAST_PATH    (ADD,  a8r8g8b8, null,     a8r8g8b8, mmx_composite_add_8888_8888       ),
diff --git a/test/lowlevel-blt-bench.c b/test/lowlevel-blt-bench.c
index 8a39a46..b44b9f8 100644
--- a/test/lowlevel-blt-bench.c
+++ b/test/lowlevel-blt-bench.c
@@ -661,6 +661,7 @@ tests_tbl[] =
     { "outrev_n_8888_1555_ca", PIXMAN_a8r8g8b8,    1, PIXMAN_OP_OUT_REV, PIXMAN_a8r8g8b8, 2, PIXMAN_a1r5g5b5 },
     { "outrev_n_8888_x888_ca", PIXMAN_a8r8g8b8,    1, PIXMAN_OP_OUT_REV, PIXMAN_a8r8g8b8, 2, PIXMAN_x8r8g8b8 },
     { "outrev_n_8888_8888_ca", PIXMAN_a8r8g8b8,    1, PIXMAN_OP_OUT_REV, PIXMAN_a8r8g8b8, 2, PIXMAN_a8r8g8b8 },
+    { "over_reverse_n_8888",   PIXMAN_a8r8g8b8,    0, PIXMAN_OP_OVER_REVERSE, PIXMAN_null, 0, PIXMAN_a8r8g8b8 },
 };
 
 int
-- 
1.7.3.4