Use 8x8 DCT transform for I8X8 prediction mode

Apply 2D-DCT transform of dimension 8x8 to encode prediction
residuals of I8X8 mode.
Brought back block type 3 probability context model for 8x8 tokens,
which is used for the coefficients of Y blocks in I8x8 modes. The
coefficient costs estimate of I8X8 mode in rate-distortion is also
changed appropriately.
Performance results:
derf:   0.246
yt:     0.114
std-hd: 0.730
hd:     0.670

Change-Id: If1d970eeb4e1827c9f0d2c5b27d33089b347ea27

1 files changed, 22 insertions, 1 deletions

diff --git a/vp8/encoder/encodeintra.c b/vp8/encoder/encodeintra.c
index 46b352ef8..01ae03a23 100644
--- a/vp8/encoder/encodeintra.c
+++ b/vp8/encoder/encodeintra.c
@@ -306,6 +306,25 @@ void vp8_encode_intra8x8(const VP8_ENCODER_RTCD *rtcd,
   }
 #endif
 
+#if CONFIG_HTRANS8X8
+  {
+    MACROBLOCKD *xd = &x->e_mbd;
+    int idx = (ib & 0x02) ? (ib + 2) : ib;
+
+    // generate residual blocks
+    vp8_subtract_4b_c(be, b, 16);
+    x->vp8_short_fdct8x8(be->src_diff, (x->block + idx)->coeff, 32);
+    x->quantize_b_8x8(x->block + idx, xd->block + idx);
+    vp8_short_idct8x8_c(xd->block[idx].dqcoeff, xd->block[ib].diff, 32);
+
+    // reconstruct submacroblock
+    for (i = 0; i < 4; i++) {
+      b = &xd->block[ib + iblock[i]];
+      vp8_recon_b_c(b->predictor, b->diff, *(b->base_dst) + b->dst,
+                    b->dst_stride);
+    }
+  }
+#else
   for (i = 0; i < 4; i++) {
     b = &x->e_mbd.block[ib + iblock[i]];
     be = &x->block[ib + iblock[i]];
@@ -314,8 +333,10 @@ void vp8_encode_intra8x8(const VP8_ENCODER_RTCD *rtcd,
     x->quantize_b(be, b);
     vp8_inverse_transform_b(IF_RTCD(&rtcd->common->idct), b, 32);
     RECON_INVOKE(&rtcd->common->recon, recon)(b->predictor,
-                                              b->diff, *(b->base_dst) + b->dst, b->dst_stride);
+                                              b->diff, *(b->base_dst) + b->dst,
+                                              b->dst_stride);
   }
+#endif
 }
 
 extern const int vp8_i8x8_block[4];