RTCD: add block subtraction functions

This commit continues the process of converting to the new RTCD
system.

Change-Id: Id8a287fdd4bd050ea4452e1582ad85520f3081be

13 files changed, 52 insertions, 310 deletions

diff --git a/vp8/common/rtcd_defs.sh b/vp8/common/rtcd_defs.sh
index 952f3b58a..1bdddc14d 100644
--- a/vp8/common/rtcd_defs.sh
+++ b/vp8/common/rtcd_defs.sh
@@ -437,5 +437,32 @@ specialize vp8_quantize_mby neon
 prototype void vp8_quantize_mbuv "struct macroblock *"
 specialize vp8_quantize_mbuv neon
 
+#
+# Block subtraction
+#
+prototype int vp8_block_error "short *coeff, short *dqcoeff"
+specialize vp8_block_error mmx sse2
+vp8_block_error_sse2=vp8_block_error_xmm
+
+prototype int vp8_mbblock_error "struct macroblock *mb, int dc"
+specialize vp8_mbblock_error mmx sse2
+vp8_mbblock_error_sse2=vp8_mbblock_error_xmm
+
+prototype int vp8_mbuverror "struct macroblock *mb"
+specialize vp8_mbuverror mmx sse2
+vp8_mbuverror_sse2=vp8_mbuverror_xmm
+
+prototype void vp8_subtract_b "struct block *be, struct blockd *bd, int pitch"
+specialize vp8_subtract_b mmx sse2 media neon
+vp8_subtract_b_media=vp8_subtract_b_armv6
+
+prototype void vp8_subtract_mby "short *diff, unsigned char *src, int src_stride, unsigned char *pred, int pred_stride"
+specialize vp8_subtract_mby mmx sse2 media neon
+vp8_subtract_mby_media=vp8_subtract_mby_armv6
+
+prototype void vp8_subtract_mbuv "short *diff, unsigned char *usrc, unsigned char *vsrc, int src_stride, unsigned char *upred, unsigned char *vpred, int pred_stride"
+specialize vp8_subtract_mbuv mmx sse2 media neon
+vp8_subtract_mbuv_media=vp8_subtract_mbuv_armv6
+
 # End of encoder only functions
 fi
diff --git a/vp8/encoder/arm/arm_csystemdependent.c b/vp8/encoder/arm/arm_csystemdependent.c
index 2d7c959a6..963b9fd08 100644
--- a/vp8/encoder/arm/arm_csystemdependent.c
+++ b/vp8/encoder/arm/arm_csystemdependent.c
@@ -29,29 +29,6 @@ void vp8_arch_arm_encoder_init(VP8_COMP *cpi)
     }
 #endif
 
-#if HAVE_MEDIA
-    if (flags & HAS_MEDIA)
-    {
-        /*cpi->rtcd.encodemb.berr                  = vp8_block_error_c;
-        cpi->rtcd.encodemb.mberr                 = vp8_mbblock_error_c;
-        cpi->rtcd.encodemb.mbuverr               = vp8_mbuverror_c;*/
-        cpi->rtcd.encodemb.subb                  = vp8_subtract_b_armv6;
-        cpi->rtcd.encodemb.submby                = vp8_subtract_mby_armv6;
-        cpi->rtcd.encodemb.submbuv               = vp8_subtract_mbuv_armv6;
-    }
-#endif /* HAVE_MEDIA */
-
-#if HAVE_NEON
-    if (flags & HAS_NEON)
-    {
-        /*cpi->rtcd.encodemb.berr                  = vp8_block_error_c;
-        cpi->rtcd.encodemb.mberr                 = vp8_mbblock_error_c;
-        cpi->rtcd.encodemb.mbuverr               = vp8_mbuverror_c;*/
-        cpi->rtcd.encodemb.subb                  = vp8_subtract_b_neon;
-        cpi->rtcd.encodemb.submby                = vp8_subtract_mby_neon;
-        cpi->rtcd.encodemb.submbuv               = vp8_subtract_mbuv_neon;
-    }
-#endif /* HAVE_NEON */
 #endif /* CONFIG_RUNTIME_CPU_DETECT */
 
 #if HAVE_NEON
diff --git a/vp8/encoder/arm/encodemb_arm.h b/vp8/encoder/arm/encodemb_arm.h
deleted file mode 100644
index 55ffdb32e..000000000
--- a/vp8/encoder/arm/encodemb_arm.h
+++ /dev/null
@@ -1,64 +0,0 @@
-/*
- *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-
-#ifndef ENCODEMB_ARM_H
-#define ENCODEMB_ARM_H
-
-#if HAVE_MEDIA
-extern prototype_subb(vp8_subtract_b_armv6);
-extern prototype_submby(vp8_subtract_mby_armv6);
-extern prototype_submbuv(vp8_subtract_mbuv_armv6);
-
-#if !CONFIG_RUNTIME_CPU_DETECT
-#undef  vp8_encodemb_subb
-#define vp8_encodemb_subb vp8_subtract_b_armv6
-
-#undef  vp8_encodemb_submby
-#define vp8_encodemb_submby vp8_subtract_mby_armv6
-
-#undef  vp8_encodemb_submbuv
-#define vp8_encodemb_submbuv vp8_subtract_mbuv_armv6
-#endif
-
-#endif /* HAVE_MEDIA */
-
-#if HAVE_NEON
-//extern prototype_berr(vp8_block_error_c);
-//extern prototype_mberr(vp8_mbblock_error_c);
-//extern prototype_mbuverr(vp8_mbuverror_c);
-
-extern prototype_subb(vp8_subtract_b_neon);
-extern prototype_submby(vp8_subtract_mby_neon);
-extern prototype_submbuv(vp8_subtract_mbuv_neon);
-
-//#undef  vp8_encodemb_berr
-//#define vp8_encodemb_berr vp8_block_error_c
-
-//#undef  vp8_encodemb_mberr
-//#define vp8_encodemb_mberr vp8_mbblock_error_c
-
-//#undef  vp8_encodemb_mbuverr
-//#define vp8_encodemb_mbuverr vp8_mbuverror_c
-
-#if !CONFIG_RUNTIME_CPU_DETECT
-#undef  vp8_encodemb_subb
-#define vp8_encodemb_subb vp8_subtract_b_neon
-
-#undef  vp8_encodemb_submby
-#define vp8_encodemb_submby vp8_subtract_mby_neon
-
-#undef  vp8_encodemb_submbuv
-#define vp8_encodemb_submbuv vp8_subtract_mbuv_neon
-#endif
-
-#endif /* HAVE_NEON */
-
-#endif
diff --git a/vp8/encoder/encodeintra.c b/vp8/encoder/encodeintra.c
index a1cc87c8b..c457906f6 100644
--- a/vp8/encoder/encodeintra.c
+++ b/vp8/encoder/encodeintra.c
@@ -67,7 +67,7 @@ void vp8_encode_intra4x4block(const VP8_ENCODER_RTCD *rtcd,
                 (*(b->base_dst) + b->dst, b->dst_stride,
                  b->bmi.as_mode, b->predictor, 16);
 
-    ENCODEMB_INVOKE(&rtcd->encodemb, subb)(be, b, 16);
+    vp8_subtract_b(be, b, 16);
 
     x->short_fdct4x4(be->src_diff, be->coeff, 32);
 
@@ -105,7 +105,7 @@ void vp8_encode_intra16x16mby(const VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x)
 
     vp8_build_intra_predictors_mby_s(&x->e_mbd);
 
-    ENCODEMB_INVOKE(&rtcd->encodemb, submby) (x->src_diff, *(b->base_src),
+    vp8_subtract_mby(x->src_diff, *(b->base_src),
         b->src_stride, xd->dst.y_buffer, xd->dst.y_stride);
 
     vp8_transform_intra_mby(x);
@@ -122,7 +122,7 @@ void vp8_encode_intra16x16mbuv(const VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x)
 
     vp8_build_intra_predictors_mbuv_s(&x->e_mbd);
 
-    ENCODEMB_INVOKE(&rtcd->encodemb, submbuv)(x->src_diff, x->src.u_buffer,
+    vp8_subtract_mbuv(x->src_diff, x->src.u_buffer,
         x->src.v_buffer, x->src.uv_stride, xd->dst.u_buffer,
         xd->dst.v_buffer, xd->dst.uv_stride);
 
diff --git a/vp8/encoder/encodemb.c b/vp8/encoder/encodemb.c
index 353eeb6fc..8e139400b 100644
--- a/vp8/encoder/encodemb.c
+++ b/vp8/encoder/encodemb.c
@@ -102,9 +102,9 @@ static void vp8_subtract_mb(const VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x)
 {
     BLOCK *b = &x->block[0];
 
-    ENCODEMB_INVOKE(&rtcd->encodemb, submby)(x->src_diff, *(b->base_src),
+    vp8_subtract_mby(x->src_diff, *(b->base_src),
         b->src_stride, x->e_mbd.dst.y_buffer, x->e_mbd.dst.y_stride);
-    ENCODEMB_INVOKE(&rtcd->encodemb, submbuv)(x->src_diff, x->src.u_buffer,
+    vp8_subtract_mbuv(x->src_diff, x->src.u_buffer,
         x->src.v_buffer, x->src.uv_stride, x->e_mbd.dst.u_buffer,
         x->e_mbd.dst.v_buffer, x->e_mbd.dst.uv_stride);
 }
@@ -643,7 +643,7 @@ void vp8_encode_inter16x16y(const VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x)
     vp8_build_inter16x16_predictors_mby(&x->e_mbd, x->e_mbd.dst.y_buffer,
                                         x->e_mbd.dst.y_stride);
 
-    ENCODEMB_INVOKE(&rtcd->encodemb, submby)(x->src_diff, *(b->base_src),
+    vp8_subtract_mby(x->src_diff, *(b->base_src),
         b->src_stride, x->e_mbd.dst.y_buffer, x->e_mbd.dst.y_stride);
 
     transform_mby(x);
diff --git a/vp8/encoder/encodemb.h b/vp8/encoder/encodemb.h
index 0fa87cf68..df3ac2f6f 100644
--- a/vp8/encoder/encodemb.h
+++ b/vp8/encoder/encodemb.h
@@ -12,88 +12,6 @@
 #ifndef __INC_ENCODEMB_H
 #define __INC_ENCODEMB_H
 
-
-#include "vpx_config.h"
-#include "block.h"
-
-#define prototype_mberr(sym) \
-    int (sym)(MACROBLOCK *mb, int dc)
-
-#define prototype_berr(sym) \
-    int (sym)(short *coeff, short *dqcoeff)
-
-#define prototype_mbuverr(sym) \
-    int (sym)(MACROBLOCK *mb)
-
-#define prototype_subb(sym) \
-    void (sym)(BLOCK *be,BLOCKD *bd, int pitch)
-
-#define prototype_submby(sym) \
-    void (sym)(short *diff, unsigned char *src, int src_stride, \
-        unsigned char *pred, int pred_stride)
-
-#define prototype_submbuv(sym) \
-    void (sym)(short *diff, unsigned char *usrc, unsigned char *vsrc,\
-               int src_stride, unsigned char *upred, unsigned char *vpred,\
-               int pred_stride)
-
-#if ARCH_X86 || ARCH_X86_64
-#include "x86/encodemb_x86.h"
-#endif
-
-#if ARCH_ARM
-#include "arm/encodemb_arm.h"
-#endif
-
-#ifndef vp8_encodemb_berr
-#define vp8_encodemb_berr vp8_block_error_c
-#endif
-extern prototype_berr(vp8_encodemb_berr);
-
-#ifndef vp8_encodemb_mberr
-#define vp8_encodemb_mberr vp8_mbblock_error_c
-#endif
-extern prototype_mberr(vp8_encodemb_mberr);
-
-#ifndef vp8_encodemb_mbuverr
-#define vp8_encodemb_mbuverr vp8_mbuverror_c
-#endif
-extern prototype_mbuverr(vp8_encodemb_mbuverr);
-
-#ifndef vp8_encodemb_subb
-#define vp8_encodemb_subb vp8_subtract_b_c
-#endif
-extern prototype_subb(vp8_encodemb_subb);
-
-#ifndef vp8_encodemb_submby
-#define vp8_encodemb_submby vp8_subtract_mby_c
-#endif
-extern prototype_submby(vp8_encodemb_submby);
-
-#ifndef vp8_encodemb_submbuv
-#define vp8_encodemb_submbuv vp8_subtract_mbuv_c
-#endif
-extern prototype_submbuv(vp8_encodemb_submbuv);
-
-
-typedef struct
-{
-    prototype_berr(*berr);
-    prototype_mberr(*mberr);
-    prototype_mbuverr(*mbuverr);
-    prototype_subb(*subb);
-    prototype_submby(*submby);
-    prototype_submbuv(*submbuv);
-} vp8_encodemb_rtcd_vtable_t;
-
-#if CONFIG_RUNTIME_CPU_DETECT
-#define ENCODEMB_INVOKE(ctx,fn) (ctx)->fn
-#else
-#define ENCODEMB_INVOKE(ctx,fn) vp8_encodemb_##fn
-#endif
-
-
-
 #include "onyx_int.h"
 struct VP8_ENCODER_RTCD;
 void vp8_encode_inter16x16(const struct VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x);
diff --git a/vp8/encoder/generic/csystemdependent.c b/vp8/encoder/generic/csystemdependent.c
index 749bcc32b..7425dc246 100644
--- a/vp8/encoder/generic/csystemdependent.c
+++ b/vp8/encoder/generic/csystemdependent.c
@@ -25,13 +25,6 @@ extern void vp8_yv12_copy_partial_frame(YV12_BUFFER_CONFIG *src_ybc,
 void vp8_cmachine_specific_config(VP8_COMP *cpi)
 {
 #if CONFIG_RUNTIME_CPU_DETECT
-    cpi->rtcd.encodemb.berr                  = vp8_block_error_c;
-    cpi->rtcd.encodemb.mberr                 = vp8_mbblock_error_c;
-    cpi->rtcd.encodemb.mbuverr               = vp8_mbuverror_c;
-    cpi->rtcd.encodemb.subb                  = vp8_subtract_b_c;
-    cpi->rtcd.encodemb.submby                = vp8_subtract_mby_c;
-    cpi->rtcd.encodemb.submbuv               = vp8_subtract_mbuv_c;
-
     cpi->rtcd.search.full_search             = vp8_full_search_sad;
     cpi->rtcd.search.refining_search         = vp8_refining_search_sad;
     cpi->rtcd.search.diamond_search          = vp8_diamond_search_sad;
diff --git a/vp8/encoder/onyx_int.h b/vp8/encoder/onyx_int.h
index 8e7f99bda..d514d75aa 100644
--- a/vp8/encoder/onyx_int.h
+++ b/vp8/encoder/onyx_int.h
@@ -223,7 +223,6 @@ typedef struct
 
 typedef struct VP8_ENCODER_RTCD
 {
-    vp8_encodemb_rtcd_vtable_t  encodemb;
     vp8_search_rtcd_vtable_t    search;
     vp8_temporal_rtcd_vtable_t  temporal;
 } VP8_ENCODER_RTCD;
diff --git a/vp8/encoder/rdopt.c b/vp8/encoder/rdopt.c
index 5ec24902a..786138f3f 100644
--- a/vp8/encoder/rdopt.c
+++ b/vp8/encoder/rdopt.c
@@ -562,8 +562,7 @@ static int vp8_rdcost_mby(MACROBLOCK *mb)
 
 static void macro_block_yrd( MACROBLOCK *mb,
                              int *Rate,
-                             int *Distortion,
-                             const vp8_encodemb_rtcd_vtable_t *rtcd)
+                             int *Distortion)
 {
     int b;
     MACROBLOCKD *const x = &mb->e_mbd;
@@ -573,7 +572,7 @@ static void macro_block_yrd( MACROBLOCK *mb,
     BLOCK *beptr;
     int d;
 
-    ENCODEMB_INVOKE(rtcd, submby)( mb->src_diff, *(mb->block[0].base_src),
+    vp8_subtract_mby( mb->src_diff, *(mb->block[0].base_src),
         mb->block[0].src_stride,  mb->e_mbd.predictor, 16);
 
     // Fdct and building the 2nd order block
@@ -597,8 +596,8 @@ static void macro_block_yrd( MACROBLOCK *mb,
     mb->quantize_b(mb_y2, x_y2);
 
     // Distortion
-    d = ENCODEMB_INVOKE(rtcd, mberr)(mb, 1) << 2;
-    d += ENCODEMB_INVOKE(rtcd, berr)(mb_y2->coeff, x_y2->dqcoeff);
+    d = vp8_mbblock_error(mb, 1) << 2;
+    d += vp8_block_error(mb_y2->coeff, x_y2->dqcoeff);
 
     *Distortion = (d >> 4);
 
@@ -654,7 +653,7 @@ static int rd_pick_intra4x4block(
         vp8_intra4x4_predict
                      (*(b->base_dst) + b->dst, b->dst_stride,
                       mode, b->predictor, 16);
-        ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), subb)(be, b, 16);
+        vp8_subtract_b(be, b, 16);
         x->short_fdct4x4(be->src_diff, be->coeff, 32);
         x->quantize_b(be, b);
 
@@ -663,7 +662,7 @@ static int rd_pick_intra4x4block(
 
         ratey = cost_coeffs(x, b, PLANE_TYPE_Y_WITH_DC, &tempa, &templ);
         rate += ratey;
-        distortion = ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), berr)(be->coeff, b->dqcoeff) >> 2;
+        distortion = vp8_block_error(be->coeff, b->dqcoeff) >> 2;
 
         this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);
 
@@ -774,7 +773,7 @@ static int rd_pick_intra16x16mby_mode(VP8_COMP *cpi,
         vp8_build_intra_predictors_mby
             (&x->e_mbd);
 
-        macro_block_yrd(x, &ratey, &distortion, IF_RTCD(&cpi->rtcd.encodemb));
+        macro_block_yrd(x, &ratey, &distortion);
         rate = ratey + x->mbmode_cost[x->e_mbd.frame_type]
                                      [x->e_mbd.mode_info_context->mbmi.mode];
 
@@ -821,7 +820,7 @@ static int rd_inter16x16_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
                             int *distortion, int fullpixel)
 {
     vp8_build_inter16x16_predictors_mbuv(&x->e_mbd);
-    ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
+    vp8_subtract_mbuv(x->src_diff,
         x->src.u_buffer, x->src.v_buffer, x->src.uv_stride,
         &x->e_mbd.predictor[256], &x->e_mbd.predictor[320], 8);
 
@@ -829,7 +828,7 @@ static int rd_inter16x16_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
     vp8_quantize_mbuv(x);
 
     *rate       = rd_cost_mbuv(x);
-    *distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
+    *distortion = vp8_mbuverror(x) / 4;
 
     return RDCOST(x->rdmult, x->rddiv, *rate, *distortion);
 }
@@ -838,7 +837,7 @@ static int rd_inter4x4_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
                           int *distortion, int fullpixel)
 {
     vp8_build_inter4x4_predictors_mbuv(&x->e_mbd);
-    ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
+    vp8_subtract_mbuv(x->src_diff,
         x->src.u_buffer, x->src.v_buffer, x->src.uv_stride,
         &x->e_mbd.predictor[256], &x->e_mbd.predictor[320], 8);
 
@@ -846,7 +845,7 @@ static int rd_inter4x4_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
     vp8_quantize_mbuv(x);
 
     *rate       = rd_cost_mbuv(x);
-    *distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
+    *distortion = vp8_mbuverror(x) / 4;
 
     return RDCOST(x->rdmult, x->rddiv, *rate, *distortion);
 }
@@ -868,7 +867,7 @@ static void rd_pick_intra_mbuv_mode(VP8_COMP *cpi, MACROBLOCK *x, int *rate, int
         x->e_mbd.mode_info_context->mbmi.uv_mode = mode;
         vp8_build_intra_predictors_mbuv
                      (&x->e_mbd);
-        ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
+        vp8_subtract_mbuv(x->src_diff,
                       x->src.u_buffer, x->src.v_buffer, x->src.uv_stride,
                       &x->e_mbd.predictor[256], &x->e_mbd.predictor[320], 8);
         vp8_transform_mbuv(x);
@@ -877,7 +876,7 @@ static void rd_pick_intra_mbuv_mode(VP8_COMP *cpi, MACROBLOCK *x, int *rate, int
         rate_to = rd_cost_mbuv(x);
         rate = rate_to + x->intra_uv_mode_cost[x->e_mbd.frame_type][x->e_mbd.mode_info_context->mbmi.uv_mode];
 
-        distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
+        distortion = vp8_mbuverror(x) / 4;
 
         this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);
 
@@ -1012,7 +1011,7 @@ static int rdcost_mbsegment_y(MACROBLOCK *mb, const int *labels,
     return cost;
 
 }
-static unsigned int vp8_encode_inter_mb_segment(MACROBLOCK *x, int const *labels, int which_label, const vp8_encodemb_rtcd_vtable_t *rtcd)
+static unsigned int vp8_encode_inter_mb_segment(MACROBLOCK *x, int const *labels, int which_label)
 {
     int i;
     unsigned int distortion = 0;
@@ -1026,14 +1025,14 @@ static unsigned int vp8_encode_inter_mb_segment(MACROBLOCK *x, int const *labels
 
 
             vp8_build_inter_predictors_b(bd, 16, x->e_mbd.subpixel_predict);
-            ENCODEMB_INVOKE(rtcd, subb)(be, bd, 16);
+            vp8_subtract_b(be, bd, 16);
             x->short_fdct4x4(be->src_diff, be->coeff, 32);
 
             // set to 0 no way to account for 2nd order DC so discount
             //be->coeff[0] = 0;
             x->quantize_b(be, bd);
 
-            distortion += ENCODEMB_INVOKE(rtcd, berr)(be->coeff, bd->dqcoeff);
+            distortion += vp8_block_error(be->coeff, bd->dqcoeff);
         }
     }
 
@@ -1275,7 +1274,7 @@ static void rd_check_segment(VP8_COMP *cpi, MACROBLOCK *x,
                 continue;
             }
 
-            distortion = vp8_encode_inter_mb_segment(x, labels, i, IF_RTCD(&cpi->rtcd.encodemb)) / 4;
+            distortion = vp8_encode_inter_mb_segment(x, labels, i) / 4;
 
             labelyrate = rdcost_mbsegment_y(x, labels, i, ta_s, tl_s);
             rate += labelyrate;
@@ -1948,7 +1947,7 @@ void vp8_rd_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
             x->e_mbd.mode_info_context->mbmi.ref_frame = INTRA_FRAME;
             vp8_build_intra_predictors_mby
                 (&x->e_mbd);
-            macro_block_yrd(x, &rate_y, &distortion, IF_RTCD(&cpi->rtcd.encodemb)) ;
+            macro_block_yrd(x, &rate_y, &distortion) ;
             rate2 += rate_y;
             distortion2 += distortion;
             rate2 += x->mbmode_cost[x->e_mbd.frame_type][x->e_mbd.mode_info_context->mbmi.mode];
@@ -2176,7 +2175,7 @@ void vp8_rd_pick_inter_mode(VP8_COMP *cpi, MACROBLOCK *x, int recon_yoffset,
             rate2 += vp8_cost_mv_ref(this_mode, mdcounts);
 
             // Y cost and distortion
-            macro_block_yrd(x, &rate_y, &distortion, IF_RTCD(&cpi->rtcd.encodemb));
+            macro_block_yrd(x, &rate_y, &distortion);
             rate2 += rate_y;
             distortion2 += distortion;
 
diff --git a/vp8/encoder/x86/encodemb_x86.h b/vp8/encoder/x86/encodemb_x86.h
deleted file mode 100644
index 69b3edd66..000000000
--- a/vp8/encoder/x86/encodemb_x86.h
+++ /dev/null
@@ -1,85 +0,0 @@
-/*
- *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-
-#ifndef ENCODEMB_X86_H
-#define ENCODEMB_X86_H
-
-
-/* Note:
- *
- * This platform is commonly built for runtime CPU detection. If you modify
- * any of the function mappings present in this file, be sure to also update
- * them in the function pointer initialization code
- */
-#if HAVE_MMX
-extern prototype_berr(vp8_block_error_mmx);
-extern prototype_mberr(vp8_mbblock_error_mmx);
-extern prototype_mbuverr(vp8_mbuverror_mmx);
-extern prototype_subb(vp8_subtract_b_mmx);
-extern prototype_submby(vp8_subtract_mby_mmx);
-extern prototype_submbuv(vp8_subtract_mbuv_mmx);
-
-
-#if !CONFIG_RUNTIME_CPU_DETECT
-#undef  vp8_encodemb_berr
-#define vp8_encodemb_berr vp8_block_error_mmx
-
-#undef  vp8_encodemb_mberr
-#define vp8_encodemb_mberr vp8_mbblock_error_mmx
-
-#undef  vp8_encodemb_mbuverr
-#define vp8_encodemb_mbuverr vp8_mbuverror_mmx
-
-#undef  vp8_encodemb_subb
-#define vp8_encodemb_subb vp8_subtract_b_mmx
-
-#undef  vp8_encodemb_submby
-#define vp8_encodemb_submby vp8_subtract_mby_mmx
-
-#undef  vp8_encodemb_submbuv
-#define vp8_encodemb_submbuv vp8_subtract_mbuv_mmx
-
-#endif
-#endif
-
-
-#if HAVE_SSE2
-extern prototype_berr(vp8_block_error_xmm);
-extern prototype_mberr(vp8_mbblock_error_xmm);
-extern prototype_mbuverr(vp8_mbuverror_xmm);
-extern prototype_subb(vp8_subtract_b_sse2);
-extern prototype_submby(vp8_subtract_mby_sse2);
-extern prototype_submbuv(vp8_subtract_mbuv_sse2);
-
-#if !CONFIG_RUNTIME_CPU_DETECT
-#undef  vp8_encodemb_berr
-#define vp8_encodemb_berr vp8_block_error_xmm
-
-#undef  vp8_encodemb_mberr
-#define vp8_encodemb_mberr vp8_mbblock_error_xmm
-
-#undef  vp8_encodemb_mbuverr
-#define vp8_encodemb_mbuverr vp8_mbuverror_xmm
-
-#undef  vp8_encodemb_subb
-#define vp8_encodemb_subb vp8_subtract_b_sse2
-
-#undef  vp8_encodemb_submby
-#define vp8_encodemb_submby vp8_subtract_mby_sse2
-
-#undef  vp8_encodemb_submbuv
-#define vp8_encodemb_submbuv vp8_subtract_mbuv_sse2
-
-#endif
-#endif
-
-
-#endif
diff --git a/vp8/encoder/x86/x86_csystemdependent.c b/vp8/encoder/x86/x86_csystemdependent.c
index 9011ddfa3..fa628fa81 100644
--- a/vp8/encoder/x86/x86_csystemdependent.c
+++ b/vp8/encoder/x86/x86_csystemdependent.c
@@ -125,29 +125,9 @@ void vp8_arch_x86_encoder_init(VP8_COMP *cpi)
      */
 
     /* Override default functions with fastest ones for this CPU. */
-#if HAVE_MMX
-    if (flags & HAS_MMX)
-    {
-        cpi->rtcd.encodemb.berr                  = vp8_block_error_mmx;
-        cpi->rtcd.encodemb.mberr                 = vp8_mbblock_error_mmx;
-        cpi->rtcd.encodemb.mbuverr               = vp8_mbuverror_mmx;
-        cpi->rtcd.encodemb.subb                  = vp8_subtract_b_mmx;
-        cpi->rtcd.encodemb.submby                = vp8_subtract_mby_mmx;
-        cpi->rtcd.encodemb.submbuv               = vp8_subtract_mbuv_mmx;
-
-        /*cpi->rtcd.quantize.fastquantb            = vp8_fast_quantize_b_mmx;*/
-    }
-#endif
-
 #if HAVE_SSE2
     if (flags & HAS_SSE2)
     {
-        cpi->rtcd.encodemb.berr                  = vp8_block_error_xmm;
-        cpi->rtcd.encodemb.mberr                 = vp8_mbblock_error_xmm;
-        cpi->rtcd.encodemb.mbuverr               = vp8_mbuverror_xmm;
-        cpi->rtcd.encodemb.subb                  = vp8_subtract_b_sse2;
-        cpi->rtcd.encodemb.submby                = vp8_subtract_mby_sse2;
-        cpi->rtcd.encodemb.submbuv               = vp8_subtract_mbuv_sse2;
 #if !(CONFIG_REALTIME_ONLY)
         cpi->rtcd.temporal.apply                 = vp8_temporal_filter_apply_sse2;
 #endif
diff --git a/vp8/vp8cx.mk b/vp8/vp8cx.mk
index 1760e4b88..febb7eff3 100644
--- a/vp8/vp8cx.mk
+++ b/vp8/vp8cx.mk
@@ -93,7 +93,6 @@ VP8_CX_SRCS_REMOVE-yes += encoder/firstpass.c
 VP8_CX_SRCS_REMOVE-yes += encoder/temporal_filter.c
 endif
 
-VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/encodemb_x86.h
 VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/mcomp_x86.h
 VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/temporal_filter_x86.h
 VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/x86_csystemdependent.c
diff --git a/vp8/vp8cx_arm.mk b/vp8/vp8cx_arm.mk
index 1fdb49377..0d37f4bfe 100644
--- a/vp8/vp8cx_arm.mk
+++ b/vp8/vp8cx_arm.mk
@@ -16,7 +16,6 @@
 VP8_CX_SRCS-$(ARCH_ARM)  += encoder/arm/arm_csystemdependent.c
 
 VP8_CX_SRCS-$(ARCH_ARM)  += encoder/arm/dct_arm.c
-VP8_CX_SRCS-$(ARCH_ARM)  += encoder/arm/encodemb_arm.h
 VP8_CX_SRCS-$(ARCH_ARM)  += encoder/arm/quantize_arm.c
 VP8_CX_SRCS-$(ARCH_ARM)  += encoder/arm/variance_arm.c