diff --git a/vp8/common/blockd.h b/vp8/common/blockd.h
index cb546e74b..9db884312 100644
--- a/vp8/common/blockd.h
+++ b/vp8/common/blockd.h
@@ -255,7 +255,7 @@ typedef struct {
   union b_mode_info bmi[16];
 } MODE_INFO;
 
-typedef struct {
+typedef struct blockd {
   short *qcoeff;
   short *dqcoeff;
   unsigned char  *predictor;
@@ -277,7 +277,7 @@ typedef struct {
   union b_mode_info bmi;
 } BLOCKD;
 
-typedef struct MacroBlockD {
+typedef struct macroblockd {
   DECLARE_ALIGNED(16, short, diff[400]);      /* from idct diff */
   DECLARE_ALIGNED(16, unsigned char,  predictor[384]);
   DECLARE_ALIGNED(16, short, qcoeff[400]);
diff --git a/vp8/common/loopfilter.h b/vp8/common/loopfilter.h
index df78b8c12..c9f4fb5a4 100644
--- a/vp8/common/loopfilter.h
+++ b/vp8/common/loopfilter.h
@@ -81,22 +81,22 @@ typedef void loop_filter_uvfunction
 
 /* assorted loopfilter functions which get used elsewhere */
 struct VP8Common;
-struct MacroBlockD;
+struct macroblockd;
 
 void vp8_loop_filter_init(struct VP8Common *cm);
 
 void vp8_loop_filter_frame_init(struct VP8Common *cm,
-                                struct MacroBlockD *mbd,
+                                struct macroblockd *mbd,
                                 int default_filt_lvl);
 
-void vp8_loop_filter_frame(struct VP8Common *cm, struct MacroBlockD *mbd);
+void vp8_loop_filter_frame(struct VP8Common *cm, struct macroblockd *mbd);
 
 void vp8_loop_filter_partial_frame(struct VP8Common *cm,
-                                   struct MacroBlockD *mbd,
+                                   struct macroblockd *mbd,
                                    int default_filt_lvl);
 
 void vp8_loop_filter_frame_yonly(struct VP8Common *cm,
-                                 struct MacroBlockD *mbd,
+                                 struct macroblockd *mbd,
                                  int default_filt_lvl);
 
 void vp8_loop_filter_update_sharpness(loop_filter_info_n *lfi,
diff --git a/vp8/common/rtcd_defs.sh b/vp8/common/rtcd_defs.sh
index d7e5b5b25..4f096b392 100644
--- a/vp8/common/rtcd_defs.sh
+++ b/vp8/common/rtcd_defs.sh
@@ -1,8 +1,15 @@
 common_forward_decls() {
 cat <<EOF
-#include "vp8/common/blockd.h"
 
 struct loop_filter_info;
+struct blockd;
+struct macroblockd;
+struct loop_filter_info;
+
+/* Encoder forward decls */
+struct block;
+struct macroblock;
+struct variance_vtable;
 
 /* Encoder forward decls */
 struct variance_vtable;
@@ -70,55 +77,55 @@ specialize vp8_recon2b sse2
 prototype void vp8_recon4b "unsigned char *pred_ptr, short *diff_ptr, unsigned char *dst_ptr, int stride"
 specialize vp8_recon4b sse2
 
-prototype void vp8_recon_mb "MACROBLOCKD *x"
+prototype void vp8_recon_mb "struct macroblockd *x"
 specialize vp8_recon_mb
 
-prototype void vp8_recon_mby "MACROBLOCKD *x"
+prototype void vp8_recon_mby "struct macroblockd *x"
 specialize vp8_recon_mby
 
-prototype void vp8_build_intra_predictors_mby_s "MACROBLOCKD *x"
+prototype void vp8_build_intra_predictors_mby_s "struct macroblockd *x"
 specialize vp8_build_intra_predictors_mby_s
 
-prototype void vp8_build_intra_predictors_sby_s "MACROBLOCKD *x"
+prototype void vp8_build_intra_predictors_sby_s "struct macroblockd *x"
 specialize vp8_build_intra_predictors_sby_s;
 
-prototype void vp8_build_intra_predictors_sbuv_s "MACROBLOCKD *x"
+prototype void vp8_build_intra_predictors_sbuv_s "struct macroblockd *x"
 specialize vp8_build_intra_predictors_sbuv_s;
 
-prototype void vp8_build_intra_predictors_mby "MACROBLOCKD *x"
+prototype void vp8_build_intra_predictors_mby "struct macroblockd *x"
 specialize vp8_build_intra_predictors_mby;
 
-prototype void vp8_build_comp_intra_predictors_mby "MACROBLOCKD *x"
+prototype void vp8_build_comp_intra_predictors_mby "struct macroblockd *x"
 specialize vp8_build_comp_intra_predictors_mby;
 
-prototype void vp8_build_intra_predictors_mby_s "MACROBLOCKD *x"
+prototype void vp8_build_intra_predictors_mby_s "struct macroblockd *x"
 specialize vp8_build_intra_predictors_mby_s;
 
-prototype void vp8_build_intra_predictors_mbuv "MACROBLOCKD *x"
+prototype void vp8_build_intra_predictors_mbuv "struct macroblockd *x"
 specialize vp8_build_intra_predictors_mbuv;
 
-prototype void vp8_build_intra_predictors_mbuv_s "MACROBLOCKD *x"
+prototype void vp8_build_intra_predictors_mbuv_s "struct macroblockd *x"
 specialize vp8_build_intra_predictors_mbuv_s;
 
-prototype void vp8_build_comp_intra_predictors_mbuv "MACROBLOCKD *x"
+prototype void vp8_build_comp_intra_predictors_mbuv "struct macroblockd *x"
 specialize vp8_build_comp_intra_predictors_mbuv;
 
-prototype void vp8_intra4x4_predict "BLOCKD *x, int b_mode, unsigned char *predictor"
+prototype void vp8_intra4x4_predict "struct blockd *x, int b_mode, unsigned char *predictor"
 specialize vp8_intra4x4_predict;
 
-prototype void vp8_comp_intra4x4_predict "BLOCKD *x, int b_mode, int second_mode, unsigned char *predictor"
+prototype void vp8_comp_intra4x4_predict "struct blockd *x, int b_mode, int second_mode, unsigned char *predictor"
 specialize vp8_comp_intra4x4_predict;
 
-prototype void vp8_intra8x8_predict "BLOCKD *x, int b_mode, unsigned char *predictor"
+prototype void vp8_intra8x8_predict "struct blockd *x, int b_mode, unsigned char *predictor"
 specialize vp8_intra8x8_predict;
 
-prototype void vp8_comp_intra8x8_predict "BLOCKD *x, int b_mode, int second_mode, unsigned char *predictor"
+prototype void vp8_comp_intra8x8_predict "struct blockd *x, int b_mode, int second_mode, unsigned char *predictor"
 specialize vp8_comp_intra8x8_predict;
 
-prototype void vp8_intra_uv4x4_predict "BLOCKD *x, int b_mode, unsigned char *predictor"
+prototype void vp8_intra_uv4x4_predict "struct blockd *x, int b_mode, unsigned char *predictor"
 specialize vp8_intra_uv4x4_predict;
 
-prototype void vp8_comp_intra_uv4x4_predict "BLOCKD *x, int b_mode, int second_mode, unsigned char *predictor"
+prototype void vp8_comp_intra_uv4x4_predict "struct blockd *x, int b_mode, int second_mode, unsigned char *predictor"
 specialize vp8_comp_intra_uv4x4_predict;
 
 #
@@ -367,6 +374,30 @@ specialize vp8_sub_pixel_mse32x32
 
 prototype unsigned int vp8_get_mb_ss "const short *"
 specialize vp8_get_mb_ss mmx sse2
+# ENCODEMB INVOKE
+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_block_error "short *coeff, short *dqcoeff, int block_size"
+specialize vp8_block_error mmx sse2
+vp8_block_error_sse2=vp8_block_error_xmm
+
+prototype void vp8_subtract_b "struct block *be, struct blockd *bd, int pitch"
+specialize vp8_subtract_b mmx sse2
+
+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
+
+prototype void vp8_subtract_mby "short *diff, unsigned char *src, unsigned char *pred, int stride"
+specialize vp8_subtract_mby mmx sse2
+
+prototype void vp8_subtract_mbuv "short *diff, unsigned char *usrc, unsigned char *vsrc, unsigned char *pred, int stride"
+specialize vp8_subtract_mbuv mmx sse2
 
 #
 # Structured Similarity (SSIM)
diff --git a/vp8/encoder/block.h b/vp8/encoder/block.h
index 48623be8c..5d0030abc 100644
--- a/vp8/encoder/block.h
+++ b/vp8/encoder/block.h
@@ -24,7 +24,7 @@ typedef struct {
   int offset;
 } search_site;
 
-typedef struct {
+typedef struct block {
   // 16 Y blocks, 4 U blocks, 4 V blocks each with 16 entries
   short *src_diff;
   short *coeff;
@@ -82,7 +82,7 @@ typedef struct {
   int64_t txfm_rd_diff[NB_TXFM_MODES];
 } PICK_MODE_CONTEXT;
 
-typedef struct {
+typedef struct macroblock {
   DECLARE_ALIGNED(16, short, src_diff[400]);  // 16x16 Y 8x8 U 8x8 V 4x4 2nd Y
   DECLARE_ALIGNED(16, short, coeff[400]);     // 16x16 Y 8x8 U 8x8 V 4x4 2nd Y
   DECLARE_ALIGNED(16, unsigned char, thismb[256]);    // 16x16 Y
diff --git a/vp8/encoder/encodeintra.c b/vp8/encoder/encodeintra.c
index f44df22ea..e0f198734 100644
--- a/vp8/encoder/encodeintra.c
+++ b/vp8/encoder/encodeintra.c
@@ -9,6 +9,7 @@
  */
 
 #include "vpx_ports/config.h"
+#include "vpx_rtcd.h"
 #include "vp8/common/idct.h"
 #include "quantize.h"
 #include "vp8/common/reconintra.h"
@@ -18,7 +19,6 @@
 #include "dct.h"
 #include "vp8/common/g_common.h"
 #include "encodeintra.h"
-#include "vpx_rtcd.h"
 
 #if CONFIG_RUNTIME_CPU_DETECT
 #define IF_RTCD(x) (x)
@@ -70,7 +70,7 @@ void vp8_encode_intra4x4block(const VP8_ENCODER_RTCD *rtcd,
   }
 #endif
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, subb)(be, b, 16);
+  vp8_subtract_b(be, b, 16);
 
   tx_type = get_tx_type(&x->e_mbd, b);
   if (tx_type != DCT_DCT) {
@@ -109,8 +109,7 @@ void vp8_encode_intra16x16mby(const VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x) {
     vp8_build_comp_intra_predictors_mby(xd);
 #endif
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, submby)(x->src_diff, *(b->base_src),
-                                           xd->predictor, b->src_stride);
+  vp8_subtract_mby(x->src_diff, *(b->base_src), xd->predictor, b->src_stride);
 
   if (tx_size == TX_16X16) {
     BLOCKD  *bd = &xd->block[0];
@@ -159,9 +158,9 @@ void vp8_encode_intra16x16mbuv(const VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x) {
   }
 #endif
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, submbuv)(x->src_diff,
-                                            x->src.u_buffer, x->src.v_buffer,
-                                            xd->predictor, x->src.uv_stride);
+  vp8_subtract_mbuv(x->src_diff, x->src.u_buffer, x->src.v_buffer,
+                    xd->predictor, x->src.uv_stride);
+
   if (tx_size == TX_4X4) {
     vp8_transform_mbuv_4x4(x);
     vp8_quantize_mbuv_4x4(x);
@@ -221,7 +220,7 @@ void vp8_encode_intra8x8(const VP8_ENCODER_RTCD *rtcd,
     for (i = 0; i < 4; i++) {
       b = &xd->block[ib + iblock[i]];
       be = &x->block[ib + iblock[i]];
-      ENCODEMB_INVOKE(&rtcd->encodemb, subb)(be, b, 16);
+      vp8_subtract_b(be, b, 16);
       x->vp8_short_fdct4x4(be->src_diff, be->coeff, 32);
       x->quantize_b_4x4(be, b);
       vp8_inverse_transform_b_4x4(IF_RTCD(&rtcd->common->idct), b, 32);
@@ -261,7 +260,7 @@ void vp8_encode_intra_uv4x4(const VP8_ENCODER_RTCD *rtcd,
   }
 #endif
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, subb)(be, b, 8);
+  vp8_subtract_b(be, b, 8);
 
   x->vp8_short_fdct4x4(be->src_diff, be->coeff, 16);
   x->quantize_b_4x4(be, b);
diff --git a/vp8/encoder/encodemb.c b/vp8/encoder/encodemb.c
index d3bd0f1dd..aaf0819b8 100644
--- a/vp8/encoder/encodemb.c
+++ b/vp8/encoder/encodemb.c
@@ -19,6 +19,7 @@
 #include "vpx_mem/vpx_mem.h"
 #include "rdopt.h"
 #include "vp8/common/systemdependent.h"
+#include "vpx_rtcd.h"
 
 #if CONFIG_RUNTIME_CPU_DETECT
 #define IF_RTCD(x) (x)
@@ -126,8 +127,10 @@ void vp8_subtract_mby_c(short *diff, unsigned char *src,
 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), x->e_mbd.predictor, b->src_stride);
-  ENCODEMB_INVOKE(&rtcd->encodemb, submbuv)(x->src_diff, x->src.u_buffer, x->src.v_buffer, x->e_mbd.predictor, x->src.uv_stride);
+  vp8_subtract_mby(x->src_diff, *(b->base_src), x->e_mbd.predictor,
+                   b->src_stride);
+  vp8_subtract_mbuv(x->src_diff, x->src.u_buffer, x->src.v_buffer,
+                    x->e_mbd.predictor, x->src.uv_stride);
 }
 
 static void build_dcblock_4x4(MACROBLOCK *x) {
@@ -942,8 +945,7 @@ void vp8_encode_inter16x16y(const VP8_ENCODER_RTCD *rtcd, MACROBLOCK *x) {
 
   vp8_build_1st_inter16x16_predictors_mby(xd, xd->predictor, 16, 0);
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, submby)(x->src_diff, *(b->base_src),
-                                           xd->predictor, b->src_stride);
+  vp8_subtract_mby(x->src_diff, *(b->base_src), xd->predictor, b->src_stride);
 
   vp8_transform_mby_4x4(x);
   vp8_quantize_mby_4x4(x);
diff --git a/vp8/encoder/encodemb.h b/vp8/encoder/encodemb.h
index 6c28ea025..6c2300e34 100644
--- a/vp8/encoder/encodemb.h
+++ b/vp8/encoder/encodemb.h
@@ -15,73 +15,6 @@
 #include "vpx_ports/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, int block_size)
-
-#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, unsigned char *pred, int stride)
-
-#define prototype_submbuv(sym) \
-  void (sym)(short *diff, unsigned char *usrc, unsigned char *vsrc,\
-             unsigned char *pred, int 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;
-
 typedef struct {
   MB_PREDICTION_MODE mode;
   MV_REFERENCE_FRAME ref_frame;
diff --git a/vp8/encoder/generic/csystemdependent.c b/vp8/encoder/generic/csystemdependent.c
index 44e83fdc7..71f8780d4 100644
--- a/vp8/encoder/generic/csystemdependent.c
+++ b/vp8/encoder/generic/csystemdependent.c
@@ -32,14 +32,6 @@ void vp8_cmachine_specific_config(VP8_COMP *cpi) {
   cpi->rtcd.fdct.fast4x4                   = vp8_short_fdct4x4_c;
   cpi->rtcd.fdct.fast8x4                   = vp8_short_fdct8x4_c;
   cpi->rtcd.fdct.walsh_short4x4            = vp8_short_walsh4x4_c;
-
-  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 ab6802509..010611cc2 100644
--- a/vp8/encoder/onyx_int.h
+++ b/vp8/encoder/onyx_int.h
@@ -351,7 +351,6 @@ typedef struct {
 typedef struct VP8_ENCODER_RTCD {
   VP8_COMMON_RTCD            *common;
   vp8_fdct_rtcd_vtable_t      fdct;
-  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 b2207cb1f..f68bf0219 100644
--- a/vp8/encoder/rdopt.c
+++ b/vp8/encoder/rdopt.c
@@ -42,7 +42,7 @@
 #include "vp8/common/seg_common.h"
 #include "vp8/common/pred_common.h"
 #include "vp8/common/entropy.h"
-
+#include "vpx_rtcd.h"
 #if CONFIG_NEWBESTREFMV
 #include "vp8/common/mvref_common.h"
 #endif
@@ -710,11 +710,8 @@ static void macro_block_yrd_4x4(MACROBLOCK *mb,
   BLOCK *beptr;
   int d;
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, submby)(
-    mb->src_diff,
-    *(mb->block[0].base_src),
-    xd->predictor,
-    mb->block[0].src_stride);
+  vp8_subtract_mby(mb->src_diff, *(mb->block[0].base_src), xd->predictor,
+                   mb->block[0].src_stride);
 
   // Fdct and building the 2nd order block
   for (beptr = mb->block; beptr < mb->block + 16; beptr += 2) {
@@ -735,9 +732,9 @@ static void macro_block_yrd_4x4(MACROBLOCK *mb,
   mb->quantize_b_4x4(mb_y2, x_y2);
 
   // Distortion
-  d = ENCODEMB_INVOKE(&rtcd->encodemb, mberr)(mb, 1);
+  vp8_mbblock_error(mb, 1);
 
-  d += ENCODEMB_INVOKE(&rtcd->encodemb, berr)(mb_y2->coeff, x_y2->dqcoeff, 16);
+  d += vp8_block_error(mb_y2->coeff, x_y2->dqcoeff, 16);
 
   *Distortion = (d >> 2);
   // rate
@@ -784,11 +781,8 @@ static void macro_block_yrd_8x8(MACROBLOCK *mb,
   BLOCKD *const x_y2  = xd->block + 24;
   int d;
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, submby)(
-    mb->src_diff,
-    *(mb->block[0].base_src),
-    xd->predictor,
-    mb->block[0].src_stride);
+  vp8_subtract_mby(mb->src_diff, *(mb->block[0].base_src), xd->predictor,
+                   mb->block[0].src_stride);
 
   vp8_transform_mby_8x8(mb);
   vp8_quantize_mby_8x8(mb);
@@ -803,8 +797,8 @@ static void macro_block_yrd_8x8(MACROBLOCK *mb,
   xd->dqcoeff[128] = 0;
   xd->dqcoeff[192] = 0;
 
-  d = ENCODEMB_INVOKE(&rtcd->encodemb, mberr)(mb, 0);
-  d += ENCODEMB_INVOKE(&rtcd->encodemb, berr)(mb_y2->coeff, x_y2->dqcoeff, 16);
+  d = vp8_mbblock_error(mb, 0);
+  d += vp8_block_error(mb_y2->coeff, x_y2->dqcoeff, 16);
 
   *Distortion = (d >> 2);
   // rate
@@ -836,11 +830,8 @@ static void macro_block_yrd_16x16(MACROBLOCK *mb, int *Rate, int *Distortion,
   BLOCK  *be = &mb->block[0];
   TX_TYPE tx_type;
 
-  ENCODEMB_INVOKE(&rtcd->encodemb, submby)(
-    mb->src_diff,
-    *(mb->block[0].base_src),
-    mb->e_mbd.predictor,
-    mb->block[0].src_stride);
+  vp8_subtract_mby(mb->src_diff, *(mb->block[0].base_src), mb->e_mbd.predictor,
+                   mb->block[0].src_stride);
 
   tx_type = get_tx_type_16x16(xd, b);
   if (tx_type != DCT_DCT) {
@@ -855,7 +846,7 @@ static void macro_block_yrd_16x16(MACROBLOCK *mb, int *Rate, int *Distortion,
   if (mb->e_mbd.mode_info_context->mbmi.mode < I8X8_PRED)
     vp8_optimize_mby_16x16(mb, rtcd);
 
-  d = ENCODEMB_INVOKE(&rtcd->encodemb, mberr)(mb, 0);
+  d = vp8_mbblock_error(mb, 0);
 
   *Distortion = (d >> 2);
   // rate
@@ -1029,8 +1020,8 @@ static void super_block_yrd_8x8(MACROBLOCK *x,
     xd->dqcoeff[128] = 0;
     xd->dqcoeff[192] = 0;
 
-    d += ENCODEMB_INVOKE(&rtcd->encodemb, mberr)(x, 0);
-    d += ENCODEMB_INVOKE(&rtcd->encodemb, berr)(by2->coeff, bdy2->dqcoeff, 16);
+    d += vp8_mbblock_error(x, 0);
+    d += vp8_block_error(by2->coeff, bdy2->dqcoeff, 16);
     xd->above_context = ta + x_idx;
     xd->left_context = tl + y_idx;
     r += vp8_rdcost_mby_8x8(x, 0);
@@ -1121,7 +1112,7 @@ static int64_t rd_pick_intra4x4block(VP8_COMP *cpi, MACROBLOCK *x, BLOCK *be,
         rate += bmode_costs[mode2];
       }
 #endif
-      ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), subb)(be, b, 16);
+      vp8_subtract_b(be, b, 16);
 
       b->bmi.as_mode.first = mode;
       tx_type = get_tx_type_4x4(xd, b);
@@ -1138,8 +1129,7 @@ static int64_t rd_pick_intra4x4block(VP8_COMP *cpi, MACROBLOCK *x, BLOCK *be,
 
       ratey = cost_coeffs(x, b, PLANE_TYPE_Y_WITH_DC, &tempa, &templ, TX_4X4);
       rate += ratey;
-      distortion = ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), berr)(
-          be->coeff, b->dqcoeff, 16) >> 2;
+      distortion = vp8_block_error(be->coeff, b->dqcoeff, 16) >> 2;
 
       this_rd = RDCOST(x->rdmult, x->rddiv, rate, distortion);
 
@@ -1621,17 +1611,14 @@ static int rd_cost_mbuv(MACROBLOCK *mb) {
 
 static int64_t rd_inter16x16_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
                                 int *distortion, int fullpixel, int *skip) {
-  ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
-                                                         x->src.u_buffer,
-                                                         x->src.v_buffer,
-                                                         x->e_mbd.predictor,
-                                                         x->src.uv_stride);
+  vp8_subtract_mbuv(x->src_diff, x->src.u_buffer, x->src.v_buffer,
+                    x->e_mbd.predictor, x->src.uv_stride);
 
   vp8_transform_mbuv_4x4(x);
   vp8_quantize_mbuv_4x4(x);
 
   *rate       = rd_cost_mbuv(x);
-  *distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
+  *distortion = vp8_mbuverror(x) / 4;
   *skip       = mbuv_is_skippable_4x4(&x->e_mbd);
 
   return RDCOST(x->rdmult, x->rddiv, *rate, *distortion);
@@ -1696,7 +1683,7 @@ static int64_t rd_inter32x32_uv_8x8(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
     xd->above_context = ta + x_idx;
     xd->left_context = tl + y_idx;
     r += rd_cost_mbuv_8x8(x, 0);
-    d += ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
+    d += vp8_mbuverror(x) / 4;
     skippable = skippable && mbuv_is_skippable_8x8(xd);
   }
 
@@ -1714,17 +1701,14 @@ static int64_t rd_inter32x32_uv_8x8(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
 
 static int64_t rd_inter16x16_uv_8x8(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
                                     int *distortion, int fullpixel, int *skip) {
-  ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
-                                                         x->src.u_buffer,
-                                                         x->src.v_buffer,
-                                                         x->e_mbd.predictor,
-                                                         x->src.uv_stride);
+  vp8_subtract_mbuv(x->src_diff, x->src.u_buffer, x->src.v_buffer,
+                    x->e_mbd.predictor, x->src.uv_stride);
 
   vp8_transform_mbuv_8x8(x);
   vp8_quantize_mbuv_8x8(x);
 
   *rate       = rd_cost_mbuv_8x8(x, 1);
-  *distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
+  *distortion = vp8_mbuverror(x) / 4;
   *skip       = mbuv_is_skippable_8x8(&x->e_mbd);
 
   return RDCOST(x->rdmult, x->rddiv, *rate, *distortion);
@@ -1734,14 +1718,14 @@ static int64_t rd_inter16x16_uv_8x8(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
 static int64_t rd_inter4x4_uv(VP8_COMP *cpi, MACROBLOCK *x, int *rate,
                               int *distortion, int *skippable, int fullpixel) {
   vp8_build_inter4x4_predictors_mbuv(&x->e_mbd);
-  ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
-                                                         x->src.u_buffer, x->src.v_buffer, x->e_mbd.predictor, x->src.uv_stride);
+  vp8_subtract_mbuv(x->src_diff, x->src.u_buffer, x->src.v_buffer,
+                    x->e_mbd.predictor, x->src.uv_stride);
 
   vp8_transform_mbuv_4x4(x);
   vp8_quantize_mbuv_4x4(x);
 
   *rate       = rd_cost_mbuv(x);
-  *distortion = ENCODEMB_INVOKE(&cpi->rtcd.encodemb, mbuverr)(x) / 4;
+  *distortion = vp8_mbuverror(x) / 4;
   *skippable  = mbuv_is_skippable_4x4(&x->e_mbd);
 
   return RDCOST(x->rdmult, x->rddiv, *rate, *distortion);
@@ -1786,9 +1770,8 @@ static void rd_pick_intra_mbuv_mode(VP8_COMP *cpi,
       }
 #endif
 
-      ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
-                                                             x->src.u_buffer, x->src.v_buffer, x->e_mbd.predictor,
-                                                             x->src.uv_stride);
+      vp8_subtract_mbuv(x->src_diff, x->src.u_buffer, x->src.v_buffer,
+                        x->e_mbd.predictor, x->src.uv_stride);
       vp8_transform_mbuv_4x4(x);
       vp8_quantize_mbuv_4x4(x);
 
@@ -1796,7 +1779,7 @@ static void rd_pick_intra_mbuv_mode(VP8_COMP *cpi,
       rate = rate_to
              + x->intra_uv_mode_cost[x->e_mbd.frame_type][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);
 
@@ -1845,9 +1828,8 @@ static void rd_pick_intra_mbuv_mode_8x8(VP8_COMP *cpi,
 
     mbmi->uv_mode = mode;
     vp8_build_intra_predictors_mbuv(&x->e_mbd);
-    ENCODEMB_INVOKE(IF_RTCD(&cpi->rtcd.encodemb), submbuv)(x->src_diff,
-                                                           x->src.u_buffer, x->src.v_buffer, x->e_mbd.predictor,
-                                                           x->src.uv_stride);
+    vp8_subtract_mbuv(x->src_diff, x->src.u_buffer, x->src.v_buffer,
+                      x->e_mbd.predictor, x->src.uv_stride);
     vp8_transform_mbuv_8x8(x);
 
     vp8_quantize_mbuv_8x8(x);
@@ -1855,7 +1837,7 @@ static void rd_pick_intra_mbuv_mode_8x8(VP8_COMP *cpi,
     rate_to = rd_cost_mbuv_8x8(x, 1);
     rate = rate_to + x->intra_uv_mode_cost[x->e_mbd.frame_type][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);
 
     if (this_rd < best_rd) {
@@ -1905,7 +1887,7 @@ static void super_block_uvrd_8x8(MACROBLOCK *x,
     vp8_quantize_mbuv_8x8(x);
     s &= mbuv_is_skippable_8x8(xd);
 
-    d += ENCODEMB_INVOKE(&rtcd->encodemb, mbuverr)(x) >> 2;
+    d += vp8_mbuverror(x) >> 2;
     xd->above_context = ta + x_idx;
     xd->left_context = tl + y_idx;
     r += rd_cost_mbuv_8x8(x, 0);
@@ -2115,10 +2097,10 @@ static int64_t encode_inter_mb_segment(MACROBLOCK *x,
       vp8_build_inter_predictors_b(bd, 16, xd->subpixel_predict);
       if (xd->mode_info_context->mbmi.second_ref_frame)
         vp8_build_2nd_inter_predictors_b(bd, 16, xd->subpixel_predict_avg);
-      ENCODEMB_INVOKE(&rtcd->encodemb, subb)(be, bd, 16);
+      vp8_subtract_b(be, bd, 16);
       x->vp8_short_fdct4x4(be->src_diff, be->coeff, 32);
       x->quantize_b_4x4(be, bd);
-      thisdistortion = vp8_block_error_c(be->coeff, bd->dqcoeff, 16);
+      thisdistortion = vp8_block_error(be->coeff, bd->dqcoeff, 16);
       *distortion += thisdistortion;
       *labelyrate += cost_coeffs(x, bd, PLANE_TYPE_Y_WITH_DC,
                                  ta + vp8_block2above[i],
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 a169b493e..053601939 100644
--- a/vp8/encoder/x86/x86_csystemdependent.c
+++ b/vp8/encoder/x86/x86_csystemdependent.c
@@ -88,25 +88,8 @@ 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;
-  }
-#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;
     cpi->rtcd.temporal.apply                 = vp8_temporal_filter_apply_sse2;
 
   }
diff --git a/vp8/vp8cx.mk b/vp8/vp8cx.mk
index 6d2f18080..bbafcdb05 100644
--- a/vp8/vp8cx.mk
+++ b/vp8/vp8cx.mk
@@ -89,7 +89,6 @@ VP8_CX_SRCS-yes += encoder/mbgraph.c
 VP8_CX_SRCS-yes += encoder/mbgraph.h
 
 
-VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/encodemb_x86.h
 VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/dct_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/quantize_x86.h