Merge vp9-preview changes into experimental branch

Incorportate vp9-preview changes by merging master branch into experimental.

Conflicts:
	test/test.mk
	vp9/common/vp9_filter.c
	vp9/common/vp9_idctllm.c
	vp9/common/vp9_invtrans.h
	vp9/common/vp9_mbpitch.c
	vp9/common/vp9_rtcd_defs.sh
	vp9/common/vp9_systemdependent.h
	vp9/common/vp9_type_aliases.h
	vp9/common/x86/vp9_asm_stubs.c
	vp9/common/x86/vp9_subpixel_mmx.asm
	vp9/decoder/vp9_decodframe.c
	vp9/decoder/vp9_dequantize.c
	vp9/decoder/vp9_dequantize.h
	vp9/decoder/vp9_onyxd_int.h
	vp9/encoder/vp9_bitstream.c
	vp9/encoder/vp9_encodeframe.c
	vp9/encoder/vp9_rdopt.c

Change-Id: I17f51c3666d1b59cf1a699f87607cbc5d30a87c5

8 files changed, 93 insertions, 98 deletions

diff --git a/vp9/decoder/vp9_dboolhuff.h b/vp9/decoder/vp9_dboolhuff.h
index c8c5c3b01..5afdd67c8 100644
--- a/vp9/decoder/vp9_dboolhuff.h
+++ b/vp9/decoder/vp9_dboolhuff.h
@@ -13,7 +13,7 @@
 
 #include <stddef.h>
 #include <limits.h>
-#include "vpx_ports/config.h"
+#include "./vpx_config.h"
 #include "vpx_ports/mem.h"
 #include "vpx/vpx_integer.h"
 
diff --git a/vp9/decoder/vp9_decodframe.c b/vp9/decoder/vp9_decodframe.c
index 36eadc482..812bf10fc 100644
--- a/vp9/decoder/vp9_decodframe.c
+++ b/vp9/decoder/vp9_decodframe.c
@@ -264,7 +264,8 @@ static void decode_16x16(VP9D_COMP *pbi, MACROBLOCKD *xd,
   if (tx_type != DCT_DCT) {
     vp9_ht_dequant_idct_add_16x16_c(tx_type, xd->qcoeff,
                                     xd->block[0].dequant, xd->predictor,
-                                    xd->dst.y_buffer, 16, xd->dst.y_stride);
+                                    xd->dst.y_buffer, 16, xd->dst.y_stride,
+                                    xd->eobs[0]);
   } else {
     vp9_dequant_idct_add_16x16(xd->qcoeff, xd->block[0].dequant,
                                xd->predictor, xd->dst.y_buffer,
@@ -310,7 +311,8 @@ static void decode_8x8(VP9D_COMP *pbi, MACROBLOCKD *xd,
       }
       tx_type = get_tx_type_8x8(xd, &xd->block[ib]);
       if (tx_type != DCT_DCT) {
-        vp9_ht_dequant_idct_add_8x8_c(tx_type, q, dq, pre, dst, 16, stride);
+        vp9_ht_dequant_idct_add_8x8_c(tx_type, q, dq, pre, dst, 16, stride,
+                                      xd->eobs[idx]);
       } else {
         vp9_dequant_idct_add_8x8_c(q, dq, pre, dst, 16, stride,
                                    0, xd->eobs[idx]);
@@ -409,7 +411,7 @@ static void decode_4x4(VP9D_COMP *pbi, MACROBLOCKD *xd,
           vp9_ht_dequant_idct_add_c(tx_type, b->qcoeff,
                                     b->dequant, b->predictor,
                                     *(b->base_dst) + b->dst, 16,
-                                    b->dst_stride);
+                                    b->dst_stride, b->eob);
         } else {
           vp9_dequant_idct_add(b->qcoeff, b->dequant, b->predictor,
                                *(b->base_dst) + b->dst, 16, b->dst_stride);
@@ -454,7 +456,8 @@ static void decode_4x4(VP9D_COMP *pbi, MACROBLOCKD *xd,
       if (tx_type != DCT_DCT) {
         vp9_ht_dequant_idct_add_c(tx_type, b->qcoeff,
                                   b->dequant, b->predictor,
-                                  *(b->base_dst) + b->dst, 16, b->dst_stride);
+                                  *(b->base_dst) + b->dst, 16, b->dst_stride,
+                                  b->eob);
       } else {
         vp9_dequant_idct_add(b->qcoeff, b->dequant, b->predictor,
                              *(b->base_dst) + b->dst, 16, b->dst_stride);
@@ -516,7 +519,7 @@ static void decode_4x4(VP9D_COMP *pbi, MACROBLOCKD *xd,
           vp9_ht_dequant_idct_add_c(tx_type, b->qcoeff,
                                     b->dequant, b->predictor,
                                     *(b->base_dst) + b->dst, 16,
-                                    b->dst_stride);
+                                    b->dst_stride, b->eob);
         } else {
           vp9_dequant_idct_add(b->qcoeff, b->dequant, b->predictor,
                                *(b->base_dst) + b->dst, 16, b->dst_stride);
@@ -570,7 +573,7 @@ static void decode_16x16_sb(VP9D_COMP *pbi, MACROBLOCKD *xd,
         tx_type, xd->qcoeff, xd->block[0].dequant,
         xd->dst.y_buffer + y_idx * 16 * xd->dst.y_stride + x_idx * 16,
         xd->dst.y_buffer + y_idx * 16 * xd->dst.y_stride + x_idx * 16,
-        xd->dst.y_stride, xd->dst.y_stride);
+        xd->dst.y_stride, xd->dst.y_stride, xd->block[0].eob);
   } else {
     vp9_dequant_idct_add_16x16(
         xd->qcoeff, xd->block[0].dequant,
@@ -609,7 +612,7 @@ static void decode_8x8_sb(VP9D_COMP *pbi, MACROBLOCKD *xd,
             + x_idx * 16 + (i & 1) * 8,
             xd->dst.y_buffer + (y_idx * 16 + (i / 2) * 8) * xd->dst.y_stride
             + x_idx * 16 + (i & 1) * 8,
-            stride, stride);
+            stride, stride, b->eob);
       } else {
         vp9_dequant_idct_add_8x8_c(
             q, dq,
@@ -666,7 +669,7 @@ static void decode_4x4_sb(VP9D_COMP *pbi, MACROBLOCKD *xd,
             + x_idx * 16 + (i & 3) * 4,
             xd->dst.y_buffer + (y_idx * 16 + (i / 4) * 4) * xd->dst.y_stride
             + x_idx * 16 + (i & 3) * 4,
-            xd->dst.y_stride, xd->dst.y_stride);
+            xd->dst.y_stride, xd->dst.y_stride, b->eob);
       } else {
         vp9_dequant_idct_add_c(
             b->qcoeff, b->dequant,
diff --git a/vp9/decoder/vp9_dequantize.c b/vp9/decoder/vp9_dequantize.c
index 4376dc3d3..72cd2771e 100644
--- a/vp9/decoder/vp9_dequantize.c
+++ b/vp9/decoder/vp9_dequantize.c
@@ -14,7 +14,6 @@
 #include "vpx_mem/vpx_mem.h"
 #include "vp9/decoder/vp9_onyxd_int.h"
 #include "vp9/common/vp9_common.h"
-
 static void add_residual(const int16_t *diff, const uint8_t *pred, int pitch,
                          uint8_t *dest, int stride, int width, int height) {
   int r, c;
@@ -61,7 +60,7 @@ void vp9_dequantize_b_c(BLOCKD *d) {
 void vp9_ht_dequant_idct_add_c(TX_TYPE tx_type, int16_t *input,
                                const int16_t *dq,
                                uint8_t *pred, uint8_t *dest,
-                               int pitch, int stride) {
+                               int pitch, int stride, uint16_t eobs) {
   int16_t output[16];
   int16_t *diff_ptr = output;
   int i;
@@ -70,7 +69,7 @@ void vp9_ht_dequant_idct_add_c(TX_TYPE tx_type, int16_t *input,
     input[i] = dq[i] * input[i];
   }
 
-  vp9_ihtllm(input, output, 4 << 1, tx_type, 4);
+  vp9_ihtllm(input, output, 4 << 1, tx_type, 4, eobs);
 
   vpx_memset(input, 0, 32);
 
@@ -80,21 +79,25 @@ void vp9_ht_dequant_idct_add_c(TX_TYPE tx_type, int16_t *input,
 void vp9_ht_dequant_idct_add_8x8_c(TX_TYPE tx_type, int16_t *input,
                                    const int16_t *dq,
                                    uint8_t *pred, uint8_t *dest,
-                                   int pitch, int stride) {
+                                   int pitch, int stride, uint16_t eobs) {
   int16_t output[64];
   int16_t *diff_ptr = output;
   int i;
+  if (eobs == 0) {
+    /* All 0 DCT coefficient */
+    vp9_copy_mem8x8(pred, pitch, dest, stride);
+  } else if (eobs > 0) {
+    input[0] = dq[0] * input[0];
+    for (i = 1; i < 64; i++) {
+      input[i] = dq[1] * input[i];
+    }
 
-  input[0] = dq[0] * input[0];
-  for (i = 1; i < 64; i++) {
-    input[i] = dq[1] * input[i];
-  }
-
-  vp9_ihtllm(input, output, 16, tx_type, 8);
+    vp9_ihtllm(input, output, 16, tx_type, 8, eobs);
 
-  vpx_memset(input, 0, 128);
+    vpx_memset(input, 0, 128);
 
-  add_residual(diff_ptr, pred, pitch, dest, stride, 8, 8);
+    add_residual(diff_ptr, pred, pitch, dest, stride, 8, 8);
+  }
 }
 
 void vp9_dequant_idct_add_c(int16_t *input, const int16_t *dq, uint8_t *pred,
@@ -256,26 +259,31 @@ void vp9_dequant_idct_add_8x8_c(int16_t *input, const int16_t *dq,
 
 void vp9_ht_dequant_idct_add_16x16_c(TX_TYPE tx_type, int16_t *input,
                                      const int16_t *dq, uint8_t *pred,
-                                     uint8_t *dest, int pitch, int stride) {
+                                     uint8_t *dest, int pitch, int stride,
+                                     uint16_t eobs) {
   int16_t output[256];
   int16_t *diff_ptr = output;
   int i;
+  if (eobs == 0) {
+    /* All 0 DCT coefficient */
+    vp9_copy_mem16x16(pred, pitch, dest, stride);
+  } else if (eobs > 0) {
+    input[0]= input[0] * dq[0];
 
-  input[0]= input[0] * dq[0];
-
-  // recover quantizer for 4 4x4 blocks
-  for (i = 1; i < 256; i++)
-    input[i] = input[i] * dq[1];
+    // recover quantizer for 4 4x4 blocks
+    for (i = 1; i < 256; i++)
+      input[i] = input[i] * dq[1];
 
-  // inverse hybrid transform
-  vp9_ihtllm(input, output, 32, tx_type, 16);
+    // inverse hybrid transform
+    vp9_ihtllm(input, output, 32, tx_type, 16, eobs);
 
-  // the idct halves ( >> 1) the pitch
-  // vp9_short_idct16x16_c(input, output, 32);
+    // the idct halves ( >> 1) the pitch
+    // vp9_short_idct16x16_c(input, output, 32);
 
-  vpx_memset(input, 0, 512);
+    vpx_memset(input, 0, 512);
 
-  add_residual(diff_ptr, pred, pitch, dest, stride, 16, 16);
+    add_residual(diff_ptr, pred, pitch, dest, stride, 16, 16);
+  }
 }
 
 void vp9_dequant_idct_add_16x16_c(int16_t *input, const int16_t *dq,
diff --git a/vp9/decoder/vp9_dequantize.h b/vp9/decoder/vp9_dequantize.h
index c578608ba..bbbc173a2 100644
--- a/vp9/decoder/vp9_dequantize.h
+++ b/vp9/decoder/vp9_dequantize.h
@@ -11,108 +11,93 @@
 
 #ifndef VP9_DECODER_VP9_DEQUANTIZE_H_
 #define VP9_DECODER_VP9_DEQUANTIZE_H_
-
 #include "vp9/common/vp9_blockd.h"
 
 #if CONFIG_LOSSLESS
-extern void vp9_dequant_idct_add_lossless_c(int16_t *input,
-                                            const int16_t *dq,
-                                            uint8_t *pred,
-                                            uint8_t *output,
+extern void vp9_dequant_idct_add_lossless_c(int16_t *input, const int16_t *dq,
+                                            unsigned char *pred,
+                                            unsigned char *output,
                                             int pitch, int stride);
-extern void vp9_dequant_dc_idct_add_lossless_c(int16_t *input,
-                                               const int16_t *dq,
-                                               uint8_t *pred,
-                                               uint8_t *output,
+extern void vp9_dequant_dc_idct_add_lossless_c(int16_t *input, const int16_t *dq,
+                                               unsigned char *pred,
+                                               unsigned char *output,
                                                int pitch, int stride, int dc);
 extern void vp9_dequant_dc_idct_add_y_block_lossless_c(int16_t *q,
                                                        const int16_t *dq,
-                                                       uint8_t *pre,
-                                                       uint8_t *dst,
+                                                       unsigned char *pre,
+                                                       unsigned char *dst,
                                                        int stride,
                                                        uint16_t *eobs,
                                                        const int16_t *dc);
-extern void vp9_dequant_idct_add_y_block_lossless_c(int16_t *q,
-                                                    const int16_t *dq,
-                                                    uint8_t *pre,
-                                                    uint8_t *dst,
+extern void vp9_dequant_idct_add_y_block_lossless_c(int16_t *q, const int16_t *dq,
+                                                    unsigned char *pre,
+                                                    unsigned char *dst,
                                                     int stride,
                                                     uint16_t *eobs);
-extern void vp9_dequant_idct_add_uv_block_lossless_c(int16_t *q,
-                                                     const int16_t *dq,
-                                                     uint8_t *pre,
-                                                     uint8_t *dst_u,
-                                                     uint8_t *dst_v,
+extern void vp9_dequant_idct_add_uv_block_lossless_c(int16_t *q, const int16_t *dq,
+                                                     unsigned char *pre,
+                                                     unsigned char *dst_u,
+                                                     unsigned char *dst_v,
                                                      int stride,
                                                      uint16_t *eobs);
-#endif  // CONFIG_LOSSLESS
+#endif
 
 typedef void (*vp9_dequant_idct_add_fn_t)(int16_t *input, const int16_t *dq,
-                                          uint8_t *pred, uint8_t *output,
-                                          int pitch, int stride);
+    unsigned char *pred, unsigned char *output, int pitch, int stride);
 typedef void(*vp9_dequant_dc_idct_add_fn_t)(int16_t *input, const int16_t *dq,
-                                            uint8_t *pred, uint8_t *output,
-                                            int pitch, int stride, int dc);
+    unsigned char *pred, unsigned char *output, int pitch, int stride, int dc);
 
-typedef void(*vp9_dequant_dc_idct_add_y_block_fn_t)(int16_t *q,
-                                                    const int16_t *dq,
-                                                    uint8_t *pre, uint8_t *dst,
-                                                    int stride, uint16_t *eobs,
-                                                    const int16_t *dc);
+typedef void(*vp9_dequant_dc_idct_add_y_block_fn_t)(int16_t *q, const int16_t *dq,
+    unsigned char *pre, unsigned char *dst, int stride, uint16_t *eobs,
+    const int16_t *dc);
 typedef void(*vp9_dequant_idct_add_y_block_fn_t)(int16_t *q, const int16_t *dq,
-                                                 uint8_t *pre, uint8_t *dst,
-                                                 int stride, uint16_t *eobs);
+    unsigned char *pre, unsigned char *dst, int stride, uint16_t *eobs);
 typedef void(*vp9_dequant_idct_add_uv_block_fn_t)(int16_t *q, const int16_t *dq,
-                                                  uint8_t *pre, uint8_t *dst_u,
-                                                  uint8_t *dst_v, int stride,
-                                                  uint16_t *eobs);
+    unsigned char *pre, unsigned char *dst_u, unsigned char *dst_v, int stride,
+    uint16_t *eobs);
 
-void vp9_ht_dequant_idct_add_c(TX_TYPE tx_type, int16_t *input,
-                               const int16_t *dq,
-                               uint8_t *pred, uint8_t *dest,
-                               int pitch, int stride);
+void vp9_ht_dequant_idct_add_c(TX_TYPE tx_type, int16_t *input, const int16_t *dq,
+                                    unsigned char *pred, unsigned char *dest,
+                                    int pitch, int stride, uint16_t eobs);
 
 void vp9_ht_dequant_idct_add_8x8_c(TX_TYPE tx_type, int16_t *input,
-                                   const int16_t *dq, uint8_t *pred,
-                                   uint8_t *dest, int pitch, int stride);
+                                   const int16_t *dq, unsigned char *pred,
+                                   unsigned char *dest, int pitch, int stride,
+                                   uint16_t eobs);
 
 void vp9_ht_dequant_idct_add_16x16_c(TX_TYPE tx_type, int16_t *input,
-                                     const int16_t *dq, uint8_t *pred,
-                                     uint8_t *dest,
-                                     int pitch, int stride);
+                                     const int16_t *dq, unsigned char *pred,
+                                     unsigned char *dest,
+                                     int pitch, int stride, uint16_t eobs);
 
 #if CONFIG_SUPERBLOCKS
-void vp9_dequant_dc_idct_add_y_block_8x8_inplace_c(int16_t *q,
-                                                   const int16_t *dq,
-                                                   uint8_t *dst,
+void vp9_dequant_dc_idct_add_y_block_8x8_inplace_c(int16_t *q, const int16_t *dq,
+                                                   unsigned char *dst,
                                                    int stride,
                                                    uint16_t *eobs,
                                                    const int16_t *dc,
                                                    MACROBLOCKD *xd);
 
-void vp9_dequant_dc_idct_add_y_block_4x4_inplace_c(int16_t *q,
-                                                   const int16_t *dq,
-                                                   uint8_t *dst,
+void vp9_dequant_dc_idct_add_y_block_4x4_inplace_c(int16_t *q, const int16_t *dq,
+                                                   unsigned char *dst,
                                                    int stride,
                                                    uint16_t *eobs,
                                                    const int16_t *dc,
                                                    MACROBLOCKD *xd);
 
-void vp9_dequant_idct_add_uv_block_8x8_inplace_c(int16_t *q,
-                                                 const int16_t *dq,
-                                                 uint8_t *dstu,
-                                                 uint8_t *dstv,
+void vp9_dequant_idct_add_uv_block_8x8_inplace_c(int16_t *q, const int16_t *dq,
+                                                 unsigned char *dstu,
+                                                 unsigned char *dstv,
                                                  int stride,
                                                  uint16_t *eobs,
                                                  MACROBLOCKD *xd);
 
-void vp9_dequant_idct_add_uv_block_4x4_inplace_c(int16_t *q,
-                                                 const int16_t *dq,
-                                                 uint8_t *dstu,
-                                                 uint8_t *dstv,
+void vp9_dequant_idct_add_uv_block_4x4_inplace_c(int16_t *q, const int16_t *dq,
+                                                 unsigned char *dstu,
+                                                 unsigned char *dstv,
                                                  int stride,
                                                  uint16_t *eobs,
                                                  MACROBLOCKD *xd);
-#endif  // CONFIG_SUPERBLOCKS
+#endif
 
-#endif  // VP9_DECODER_VP9_DEQUANTIZE_H_
+#endif
diff --git a/vp9/decoder/vp9_onyxd_int.h b/vp9/decoder/vp9_onyxd_int.h
index 6b7184fbe..64975468d 100644
--- a/vp9/decoder/vp9_onyxd_int.h
+++ b/vp9/decoder/vp9_onyxd_int.h
@@ -10,8 +10,7 @@
 
 #ifndef VP9_DECODER_VP9_ONYXD_INT_H_
 #define VP9_DECODER_VP9_ONYXD_INT_H_
-
-#include "vpx_ports/config.h"
+#include "./vpx_config.h"
 #include "vp9/decoder/vp9_onyxd.h"
 #include "vp9/decoder/vp9_treereader.h"
 #include "vp9/common/vp9_onyxc_int.h"
diff --git a/vp9/decoder/x86/vp9_idct_blk_mmx.c b/vp9/decoder/x86/vp9_idct_blk_mmx.c
index df3485233..8279eaa4a 100644
--- a/vp9/decoder/x86/vp9_idct_blk_mmx.c
+++ b/vp9/decoder/x86/vp9_idct_blk_mmx.c
@@ -8,7 +8,7 @@
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
-#include "vpx_ports/config.h"
+#include "./vpx_config.h"
 #include "vp9/common/vp9_blockd.h"
 #include "vp9/decoder/vp9_dequantize.h"
 #include "vp9/decoder/x86/vp9_idct_mmx.h"
diff --git a/vp9/decoder/x86/vp9_idct_blk_sse2.c b/vp9/decoder/x86/vp9_idct_blk_sse2.c
index 6c1fd1439..badd97f73 100644
--- a/vp9/decoder/x86/vp9_idct_blk_sse2.c
+++ b/vp9/decoder/x86/vp9_idct_blk_sse2.c
@@ -8,7 +8,7 @@
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
-#include "vpx_ports/config.h"
+#include "./vpx_config.h"
 #include "vp9/common/vp9_blockd.h"
 #include "vp9/decoder/vp9_dequantize.h"
 
diff --git a/vp9/decoder/x86/vp9_x86_dsystemdependent.c b/vp9/decoder/x86/vp9_x86_dsystemdependent.c
index d1cc53fce..51ee8ec31 100644
--- a/vp9/decoder/x86/vp9_x86_dsystemdependent.c
+++ b/vp9/decoder/x86/vp9_x86_dsystemdependent.c
@@ -8,7 +8,7 @@
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
-#include "vpx_ports/config.h"
+#include "./vpx_config.h"
 #include "vpx_ports/x86.h"
 #include "vp9/decoder/vp9_onyxd_int.h"