[x265] [PATCH Only Review] ASM routine for luma filter for 5xN block size

[nabajit at multicorewareinc.com]

# HG changeset patch
# User Nabajit Deka
# Date 1381498811 -19800
#      Fri Oct 11 19:10:11 2013 +0530
# Node ID e2c8e5de13d694510c7ac44af208d659196e39af
# Parent  14473151a9e3bb8426fbb363c0cc61aa071cae0e
ASM routine for luma filter for 5xN block size.

diff -r 14473151a9e3 -r e2c8e5de13d6 source/common/x86/ipfilter8.asm
--- a/source/common/x86/ipfilter8.asm	Fri Oct 11 18:56:03 2013 +0530
+++ b/source/common/x86/ipfilter8.asm	Fri Oct 11 19:10:11 2013 +0530
@@ -130,7 +130,7 @@
     RET
 
 %endif  ; ARCH_X86_64 == 0
-
+
 SECTION_RODATA 32
 
 tab_leftmask:   db -1, -1, -1, -1, -1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
@@ -176,3 +176,51 @@
     dec         r4d
     jnz         .loop
     RET
+
+	SECTION_RODATA 32
+
+tab_Tm8:    db 0, 1, 2, 3, 4, 5, 6, 7, 1, 2, 3, 4, 5, 6, 7, 8
+            db 2, 3, 4, 5, 6, 7, 8, 9, 3, 4, 5, 6, 7, 8, 9, 10
+            db 4, 5, 6, 7, 8, 9, 10, 11, 5, 6, 7, 8, 9, 10, 11, 12
+            db 6, 7, 8, 9, 10, 11, 12, 13, 7, 8, 9, 10, 11, 12, 13, 14
+
+tab_c_512:  times 8 dw 512
+
+SECTION .text
+
+%macro FILTER_H8 3
+    movu        %1, [srcq - 3 ]
+    pshufb      %2, %1, [tab_Tm8]
+    pmaddubsw   %2, m3
+    pshufb      m7, %1, [tab_Tm8 + 16]
+    pmaddubsw   m7, m3
+    phaddw      %2, m7
+    pshufb      m7, %1, [tab_Tm8 + 32]
+    pmaddubsw   m7, m3
+    phaddw      m7, m7
+    phaddw      %2, m7
+    pmulhrsw    %2, %3
+    packuswb    %2, %2
+    movd        [r2], %2
+    pextrb      [r2 + 4], %2, 4
+%endmacro
+
+;-----------------------------------------------------------------------------
+; void filterHorizontal_p_p_8(pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int width, int height, short const *coeff)
+;-----------------------------------------------------------------------------
+INIT_XMM sse4
+cglobal filterHorizontal_p_p_8, 4, 5, 5, src, srcStride, dst, dstStride
+
+    mov       r4, r6m
+    movu      m3, [r4]
+    packsswb  m3, m3
+    mova      m2, [tab_c_512]
+    mov       r4, r5m
+
+.loop
+    FILTER_H8   m0, m1, m2
+    add         srcq, srcStrideq
+    add         dstq, dstStrideq
+    dec r4d
+    jnz .loop
+    RET