<div dir="ltr"><br><div class="gmail_extra"><br><br><div class="gmail_quote">On Tue, Oct 15, 2013 at 3:52 AM,  <span dir="ltr"><<a href="mailto:dnyaneshwar@multicorewareinc.com" target="_blank">dnyaneshwar@multicorewareinc.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"># HG changeset patch<br>
# User Dnyaneshwar Gorade <<a href="mailto:dnyaneshwar@multicorewareinc.com">dnyaneshwar@multicorewareinc.com</a>><br>
# Date 1381826441 -19800<br>
#      Tue Oct 15 14:10:41 2013 +0530<br>
# Node ID e00905521f015292f9ccc1ae83b1476c3c433310<br>
# Parent  1019dc55ec1b7f6b4ba7d15126d96167998c4031<br>
pixel-sse41.cpp: Modified sse_ss8 function using faster macro PROCESS_SSE_SS4x1.<br></blockquote><div><br></div><div>these all need to be packaged together in one patch with the change in the macro itself, otherwise the build is broken for a stretch of commits, which is always bad when you're trying to isolate a change that introduced a bug</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
diff -r 1019dc55ec1b -r e00905521f01 source/common/vec/pixel-sse41.cpp<br>
--- a/source/common/vec/pixel-sse41.cpp Tue Oct 15 14:07:36 2013 +0530<br>
+++ b/source/common/vec/pixel-sse41.cpp Tue Oct 15 14:10:41 2013 +0530<br>
@@ -5361,14 +5361,16 @@<br>
 {<br>
     int rows = ly;<br>
     __m128i sum  = _mm_setzero_si128();<br>
-    __m128i m1, n1, diff, sign1, sign2, tmp1, tmp2;<br>
+    __m128i m1, n1, diff, sign, tmp1, tmp2;<br>
<br>
     for (; rows != 0; rows--)<br>
     {<br>
         PROCESS_SSE_SS4x1(0);<br>
<br>
-        m1 = _mm_unpackhi_epi16(m1, sign1);<br>
-        n1 = _mm_unpackhi_epi16(n1, sign2);<br>
+        sign = _mm_srai_epi16(m1, 15);<br>
+        m1 = _mm_unpackhi_epi16(m1, sign);<br>
+        sign = _mm_srai_epi16(n1, 15);<br>
+        n1 = _mm_unpackhi_epi16(n1, sign);<br>
         diff = _mm_sub_epi32(m1, n1);<br>
         diff = _mm_mullo_epi32(diff, diff);<br>
         sum = _mm_add_epi32(sum, diff);<br>
_______________________________________________<br>
x265-devel mailing list<br>
<a href="mailto:x265-devel@videolan.org">x265-devel@videolan.org</a><br>
<a href="https://mailman.videolan.org/listinfo/x265-devel" target="_blank">https://mailman.videolan.org/listinfo/x265-devel</a><br>
</blockquote></div><br><br clear="all"><div><br></div>-- <br>Steve Borho
</div></div>