[x265] [PATCH] asm: ssse3 10bit code for convert_p2s[4xN]

[rajesh at multicorewareinc.com]

# HG changeset patch
# User Rajesh Paulraj<rajesh at multicorewareinc.com>
# Date 1429015245 -19800
#      Tue Apr 14 18:10:45 2015 +0530
# Node ID cabc6ea8231ed3f530d9ffa7e4908dd31a582f5e
# Parent  dd456de98c239b86e29bf349881854a699056240
asm: ssse3 10bit code for convert_p2s[4xN]

     convert_p2s[4x4](2.70x), convert_p2s[4x8](3.53x), convert_p2s[4x16](3.82x)

diff -r dd456de98c23 -r cabc6ea8231e source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Tue Apr 14 13:41:40 2015 +0800
+++ b/source/common/x86/asm-primitives.cpp	Tue Apr 14 18:10:45 2015 +0530
@@ -948,6 +948,10 @@
         p.cu[BLOCK_16x16].count_nonzero = x265_count_nonzero_16x16_ssse3;
         p.cu[BLOCK_32x32].count_nonzero = x265_count_nonzero_32x32_ssse3;
         p.frameInitLowres = x265_frame_init_lowres_core_ssse3;
+
+        p.pu[LUMA_4x4].convert_p2s = x265_filterPixelToShort_4x4_ssse3;
+        p.pu[LUMA_4x8].convert_p2s = x265_filterPixelToShort_4x8_ssse3;
+        p.pu[LUMA_4x16].convert_p2s = x265_filterPixelToShort_4x16_ssse3;
     }
     if (cpuMask & X265_CPU_SSE4)
     {
diff -r dd456de98c23 -r cabc6ea8231e source/common/x86/ipfilter16.asm
--- a/source/common/x86/ipfilter16.asm	Tue Apr 14 13:41:40 2015 +0800
+++ b/source/common/x86/ipfilter16.asm	Tue Apr 14 18:10:45 2015 +0530
@@ -117,6 +117,7 @@
 cextern pd_32
 cextern pw_pixel_max
 cextern pd_n32768
+cextern pw_2000
 
 ;------------------------------------------------------------------------------------------------------------
 ; void interp_8tap_horiz_pp_4x4(pixel *src, intptr_t srcStride, pixel *dst, intptr_t dstStride, int coeffIdx)
@@ -5525,65 +5526,42 @@
     FILTER_VER_LUMA_SS 64, 16
     FILTER_VER_LUMA_SS 16, 64
 
-;--------------------------------------------------------------------------------------------------
-; void filterConvertPelToShort(pixel *src, intptr_t srcStride, int16_t *dst, int width, int height)
-;--------------------------------------------------------------------------------------------------
-INIT_XMM sse2
-cglobal luma_p2s, 3, 7, 5
-
-    add         r1, r1
-
-    ; load width and height
-    mov         r3d, r3m
-    mov         r4d, r4m
+;-----------------------------------------------------------------------------
+; void filterPixelToShort(pixel *src, intptr_t srcStride, int16_t *dst, intptr_t dstStride)
+;-----------------------------------------------------------------------------
+%macro P2S_H_4xN 1
+INIT_XMM ssse3
+cglobal filterPixelToShort_4x%1, 3, 6, 2
+    add        r1d, r1d
+    mov        r3d, r3m
+    add        r3d, r3d
+    lea        r4, [r3 * 3]
+    lea        r5, [r1 * 3]
 
     ; load constant
-    mova        m4, [tab_c_n8192]
-
-.loopH:
-
-    xor         r5d, r5d
-.loopW:
-    lea         r6, [r0 + r5 * 2]
-
-    movu        m0, [r6]
-    psllw       m0, 4
-    paddw       m0, m4
-
-    movu        m1, [r6 + r1]
-    psllw       m1, 4
-    paddw       m1, m4
-
-    movu        m2, [r6 + r1 * 2]
-    psllw       m2, 4
-    paddw       m2, m4
-
-    lea         r6, [r6 + r1 * 2]
-    movu        m3, [r6 + r1]
-    psllw       m3, 4
-    paddw       m3, m4
-
-    add         r5, 8
-    cmp         r5, r3
-    jg          .width4
-    movu        [r2 + r5 * 2 + FENC_STRIDE * 0 - 16], m0
-    movu        [r2 + r5 * 2 + FENC_STRIDE * 2 - 16], m1
-    movu        [r2 + r5 * 2 + FENC_STRIDE * 4 - 16], m2
-    movu        [r2 + r5 * 2 + FENC_STRIDE * 6 - 16], m3
-    je          .nextH
-    jmp         .loopW
-
-.width4:
-    movh        [r2 + r5 * 2 + FENC_STRIDE * 0 - 16], m0
-    movh        [r2 + r5 * 2 + FENC_STRIDE * 2 - 16], m1
-    movh        [r2 + r5 * 2 + FENC_STRIDE * 4 - 16], m2
-    movh        [r2 + r5 * 2 + FENC_STRIDE * 6 - 16], m3
-
-.nextH:
-    lea         r0, [r0 + r1 * 4]
-    add         r2, FENC_STRIDE * 8
-
-    sub         r4d, 4
-    jnz         .loopH
-
+    mova       m1, [pw_2000]
+
+%rep %1/4
+    movh       m0, [r0]
+    movhps     m0, [r0 + r1]
+    psllw      m0, 4
+    psubw      m0, m1
+    movh       [r2 + r3 * 0], m0
+    movhps     [r2 + r3 * 1], m0
+
+    movh       m0, [r0 + r1 * 2]
+    movhps     m0, [r0 + r5]
+    psllw      m0, 4
+    psubw      m0, m1
+    movh       [r2 + r3 * 2], m0
+    movhps     [r2 + r4], m0
+
+    lea        r0, [r0 + r1 * 4]
+    lea        r2, [r2 + r3 * 4]
+%endrep
     RET
+%endmacro
+P2S_H_4xN 4
+P2S_H_4xN 8
+P2S_H_4xN 16
+P2S_H_4xN 32
diff -r dd456de98c23 -r cabc6ea8231e source/common/x86/ipfilter8.h
--- a/source/common/x86/ipfilter8.h	Tue Apr 14 13:41:40 2015 +0800
+++ b/source/common/x86/ipfilter8.h	Tue Apr 14 18:10:45 2015 +0530
@@ -289,6 +289,9 @@
     SETUP_CHROMA_420_HORIZ_FUNC_DEF(64, 16, cpu); \
     SETUP_CHROMA_420_HORIZ_FUNC_DEF(16, 64, cpu)
 
+void x265_filterPixelToShort_4x4_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride);
+void x265_filterPixelToShort_4x8_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride);
+void x265_filterPixelToShort_4x16_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride);
 
 CHROMA_420_VERT_FILTERS(_sse2);
 CHROMA_420_HORIZ_FILTERS(_sse4);