[Pixman] [PATCH 11/11] mmx: optimize add_8_8

[Matt Turner]

Loongson:
add_8_8 =  L1: 523.72  L2: 392.06  M:113.60 ( 79.78%)  HT: 58.28  VT: 55.76  R: 43.26  RT: 14.55 ( 146Kops/s)
add_8_8 =  L1: 544.04  L2: 407.85  M:114.49 ( 80.01%)  HT: 74.33  VT: 61.84  R: 52.84  RT: 18.40 ( 172Kops/s)

ARM/iwMMXt:
add_8_8 =  L1: 213.01  L2: 221.38  M:103.59 ( 62.44%)  HT: 52.84  VT: 48.04  R: 44.53  RT: 18.19 ( 131Kops/s)
add_8_8 =  L1: 285.81  L2: 217.86  M:102.16 ( 60.34%)  HT: 56.68  VT: 53.97  R: 47.76  RT: 19.64 ( 143Kops/s)

Signed-off-by: Matt Turner <mattst88 at gmail.com>
---
 pixman/pixman-mmx.c |   30 ++++++++++++++++++++----------
 1 files changed, 20 insertions(+), 10 deletions(-)

diff --git a/pixman/pixman-mmx.c b/pixman/pixman-mmx.c
index 8ad0135..a69aa5b 100644
--- a/pixman/pixman-mmx.c
+++ b/pixman/pixman-mmx.c
@@ -2858,11 +2858,16 @@ mmx_composite_add_8_8 (pixman_implementation_t *imp,
 	while (w && (unsigned long)dst & 7)
 	{
 	    s = *src;
-	    d = *dst;
-	    t = d + s;
-	    s = t | (0 - (t >> 8));
-	    *dst = s;
-
+	    if (s)
+	    {
+		if (s != 0xff)
+		{
+		    d = *dst;
+		    t = d + s;
+		    s = t | (0 - (t >> 8));
+		}
+		*dst = s;
+	    }
 	    dst++;
 	    src++;
 	    w--;
@@ -2879,11 +2884,16 @@ mmx_composite_add_8_8 (pixman_implementation_t *imp,
 	while (w)
 	{
 	    s = *src;
-	    d = *dst;
-	    t = d + s;
-	    s = t | (0 - (t >> 8));
-	    *dst = s;
-
+	    if (s)
+	    {
+		if (s != 0xff)
+		{
+		    d = *dst;
+		    t = d + s;
+		    s = t | (0 - (t >> 8));
+		}
+		*dst = s;
+	    }
 	    dst++;
 	    src++;
 	    w--;
-- 
1.7.3.4