Move x86 half-pel assembly from dsputil to hpeldsp.

libavcodec/hpeldsp.c

@@ -53,8 +53,8 @@ av_cold void ff_hpeldsp_init(HpelDSPContext* c, int flags)
     hpel_funcs(avg, [3],  2);
     hpel_funcs(avg_no_rnd,, 16);
 
-#if 0
     if (ARCH_X86)        ff_hpeldsp_init_x86   (c, flags);
+#if 0
     if (ARCH_ARM)        ff_hpeldsp_init_arm   (c, flags);
     if (HAVE_VIS)        ff_hpeldsp_init_vis   (c, flags);
     if (ARCH_ALPHA)      ff_hpeldsp_init_alpha (c, flags);

libavcodec/x86/Makefile

@@ -10,6 +10,7 @@ OBJS-$(CONFIG_H264CHROMA)              += x86/h264chroma_init.o
 OBJS-$(CONFIG_H264DSP)                 += x86/h264dsp_init.o
 OBJS-$(CONFIG_H264PRED)                += x86/h264_intrapred_init.o
 OBJS-$(CONFIG_H264QPEL)                += x86/h264_qpel.o
+OBJS-$(CONFIG_HPELDSP)                 += x86/hpeldsp_init.o
 OBJS-$(CONFIG_LPC)                     += x86/lpc.o
 OBJS-$(CONFIG_MLP_DECODER)             += x86/mlpdsp.o
 OBJS-$(CONFIG_MPEGAUDIODSP)            += x86/mpegaudiodec.o
@@ -66,7 +67,10 @@ YASM-OBJS-$(CONFIG_H264PRED)           += x86/h264_intrapred.o          \
                                           x86/h264_intrapred_10bit.o
 YASM-OBJS-$(CONFIG_H264QPEL)           += x86/h264_qpel_8bit.o          \
                                           x86/h264_qpel_10bit.o         \
-                                          x86/qpelbase.o
+                                          x86/qpelbase.o                \
+                                          x86/fpelbase.o
+YASM-OBJS-$(CONFIG_HPELDSP)            += x86/hpeldsp.o                 \
+                                          x86/fpelbase.o
 YASM-OBJS-$(CONFIG_MPEGAUDIODSP)       += x86/imdct36.o
 YASM-OBJS-$(CONFIG_PNG_DECODER)        += x86/pngdsp.o
 YASM-OBJS-$(CONFIG_PRORES_DECODER)     += x86/proresdsp.o
@@ -83,9 +87,9 @@ YASM-OBJS-$(CONFIG_VP6_DECODER)        += x86/vp56dsp.o
 YASM-OBJS-$(CONFIG_VP8_DECODER)        += x86/vp8dsp.o
 
 YASM-OBJS-$(CONFIG_DSPUTIL)            += x86/dsputil.o                 \
-                                          x86/hpeldsp.o                 \
                                           x86/mpeg4qpel.o               \
-                                          x86/qpelbase.o
+                                          x86/qpelbase.o                \
+                                          x86/fpelbase.o
 
 YASM-OBJS                              += x86/deinterlace.o             \
                                           x86/fmtconvert.o

libavcodec/x86/fpelbase.asm

+;******************************************************************************
+;* MMX optimized DSP utils
+;* Copyright (c) 2008 Loren Merritt
+;* Copyright (c) 2003-2013 Michael Niedermayer
+;* Copyright (c) 2013 Daniel Kang
+;*
+;* This file is part of FFmpeg.
+;*
+;* FFmpeg is free software; you can redistribute it and/or
+;* modify it under the terms of the GNU Lesser General Public
+;* License as published by the Free Software Foundation; either
+;* version 2.1 of the License, or (at your option) any later version.
+;*
+;* FFmpeg is distributed in the hope that it will be useful,
+;* but WITHOUT ANY WARRANTY; without even the implied warranty of
+;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+;* Lesser General Public License for more details.
+;*
+;* You should have received a copy of the GNU Lesser General Public
+;* License along with FFmpeg; if not, write to the Free Software
+;* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+;******************************************************************************
+
+%include "libavutil/x86/x86util.asm"
+
+SECTION .text
+
+INIT_MMX mmxext
+; void pixels(uint8_t *block, const uint8_t *pixels, int line_size, int h)
+%macro PIXELS48 2
+%if %2 == 4
+%define OP movh
+%else
+%define OP mova
+%endif
+cglobal %1_pixels%2, 4,5
+    movsxdifnidn r2, r2d
+    lea          r4, [r2*3]
+.loop:
+    OP           m0, [r1]
+    OP           m1, [r1+r2]
+    OP           m2, [r1+r2*2]
+    OP           m3, [r1+r4]
+    lea          r1, [r1+r2*4]
+%ifidn %1, avg
+    pavgb        m0, [r0]
+    pavgb        m1, [r0+r2]
+    pavgb        m2, [r0+r2*2]
+    pavgb        m3, [r0+r4]
+%endif
+    OP         [r0], m0
+    OP      [r0+r2], m1
+    OP    [r0+r2*2], m2
+    OP      [r0+r4], m3
+    sub         r3d, 4
+    lea          r0, [r0+r2*4]
+    jne       .loop
+    RET
+%endmacro
+
+PIXELS48 put, 4
+PIXELS48 avg, 4
+PIXELS48 put, 8
+PIXELS48 avg, 8
+
+
+INIT_XMM sse2
+; void put_pixels16_sse2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
+cglobal put_pixels16, 4,5,4
+    lea          r4, [r2*3]
+.loop:
+    movu         m0, [r1]
+    movu         m1, [r1+r2]
+    movu         m2, [r1+r2*2]
+    movu         m3, [r1+r4]
+    lea          r1, [r1+r2*4]
+    mova       [r0], m0
+    mova    [r0+r2], m1
+    mova  [r0+r2*2], m2
+    mova    [r0+r4], m3
+    sub         r3d, 4
+    lea          r0, [r0+r2*4]
+    jnz       .loop
+    REP_RET
+
+; void avg_pixels16_sse2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
+cglobal avg_pixels16, 4,5,4
+    lea          r4, [r2*3]
+.loop:
+    movu         m0, [r1]
+    movu         m1, [r1+r2]
+    movu         m2, [r1+r2*2]
+    movu         m3, [r1+r4]
+    lea          r1, [r1+r2*4]
+    pavgb        m0, [r0]
+    pavgb        m1, [r0+r2]
+    pavgb        m2, [r0+r2*2]
+    pavgb        m3, [r0+r4]
+    mova       [r0], m0
+    mova    [r0+r2], m1
+    mova  [r0+r2*2], m2
+    mova    [r0+r4], m3
+    sub         r3d, 4
+    lea          r0, [r0+r2*4]
+    jnz       .loop
+    REP_RET

libavcodec/x86/qpelbase.asm

@@ -174,83 +174,3 @@ cglobal %1_pixels16_l2, 6,6
 INIT_MMX mmxext
 PIXELS16_L2 put
 PIXELS16_L2 avg
-
-INIT_MMX mmxext
-; void pixels(uint8_t *block, const uint8_t *pixels, int line_size, int h)
-%macro PIXELS48 2
-%if %2 == 4
-%define OP movh
-%else
-%define OP mova
-%endif
-cglobal %1_pixels%2, 4,5
-    movsxdifnidn r2, r2d
-    lea          r4, [r2*3]
-.loop:
-    OP           m0, [r1]
-    OP           m1, [r1+r2]
-    OP           m2, [r1+r2*2]
-    OP           m3, [r1+r4]
-    lea          r1, [r1+r2*4]
-%ifidn %1, avg
-    pavgb        m0, [r0]
-    pavgb        m1, [r0+r2]
-    pavgb        m2, [r0+r2*2]
-    pavgb        m3, [r0+r4]
-%endif
-    OP         [r0], m0
-    OP      [r0+r2], m1
-    OP    [r0+r2*2], m2
-    OP      [r0+r4], m3
-    sub         r3d, 4
-    lea          r0, [r0+r2*4]
-    jne       .loop
-    RET
-%endmacro
-
-PIXELS48 put, 4
-PIXELS48 avg, 4
-PIXELS48 put, 8
-PIXELS48 avg, 8
-
-
-INIT_XMM sse2
-; void put_pixels16_sse2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
-cglobal put_pixels16, 4,5,4
-    lea          r4, [r2*3]
-.loop:
-    movu         m0, [r1]
-    movu         m1, [r1+r2]
-    movu         m2, [r1+r2*2]
-    movu         m3, [r1+r4]
-    lea          r1, [r1+r2*4]
-    mova       [r0], m0
-    mova    [r0+r2], m1
-    mova  [r0+r2*2], m2
-    mova    [r0+r4], m3
-    sub         r3d, 4
-    lea          r0, [r0+r2*4]
-    jnz       .loop
-    REP_RET
-
-; void avg_pixels16_sse2(uint8_t *block, const uint8_t *pixels, ptrdiff_t line_size, int h)
-cglobal avg_pixels16, 4,5,4
-    lea          r4, [r2*3]
-.loop:
-    movu         m0, [r1]
-    movu         m1, [r1+r2]
-    movu         m2, [r1+r2*2]
-    movu         m3, [r1+r4]
-    lea          r1, [r1+r2*4]
-    pavgb        m0, [r0]
-    pavgb        m1, [r0+r2]
-    pavgb        m2, [r0+r2*2]
-    pavgb        m3, [r0+r4]
-    mova       [r0], m0
-    mova    [r0+r2], m1
-    mova  [r0+r2*2], m2
-    mova    [r0+r4], m3
-    sub         r3d, 4
-    lea          r0, [r0+r2*4]
-    jnz       .loop
-    REP_RET