Change-Id: Ic40d56fb162f4e201547dfae33e62ccd9e865889

Change-Id: I0e789bdeaed60f920f7a470e56a8d4ea374233fc

Uses reduced arrays for probabilities and branch counts in the
encoder. No change in bitstream.

Change-Id: Iec605446f44db4cd325eb45fa12a3003a6ee29db

This reverts commit df037b61

Change-Id: I1a529f2590df7bc912f5035d22311268933e3dd6

Change-Id: I81c637c61ecc33cb66beb59a2a33166d66b9a0a2

1. remove prediction mode conversion
2. unified bmode, same for key and non-key frame
3. set I4X4_PRED count for pdf to 0, as I4X4_PRED is no longer
coded ever. It is determined by ref_frame and block partition

Change-Id: If5b282957c24339b241acdb9f2afef85658fe47d

Saves 1 add, 3 shifts (and a shift bias) per 1-D transform.

Change-Id: I1104bb1679fe342b2f9677df8a9cdc0cb9699e7d

This commit removes the use of bmi_ in the first-pass encoding by
forcing encode_intra4x4block_ to use DC_PRED, followed by DCT_DCT
only, as John suggested. This makes the need for bmi buffer only
up to 4 entries, instead of 16.

Change-Id: I3410007dfae789ee46a09ae20c39d3ce3c7954aa

Change-Id: I07b8c15f33e6e7c63dd0033c18c4ac5c0303cf32

Change-Id: I21e7c4ef6bc80f4b9281fc94c88fb710b1595c23

Hardware implementation needs to load coeff probs based on the
transform size. For selectable transform size, moving these bits
earlier in the bitstream adds some delay giving time to preload
the probs and speeds up the decoding process.

Change-Id: I3bfc1f662ae6f219c9286fe9ae6310c7d8a63ea7

Also do per-partition motion vector referencing in <sb8x8 partitions,
and adjust mvref finding for sub8x8 partitions.

Change-Id: Id3ed1ed4d2a8910d11d327db6cc63b8eb79f941f

Change-Id: I1a979bf74c286b157c31bab6bdcba0494acb4918

This commit fixes a transform size enc/dec mismatch issue in the
key frame coding.

Change-Id: I0c4f40464a367b33dd91ace84506650b1aec2873

This code does not seem to be necessary anymore.

For the 1080p clip used, the decoder performance improved by
~2%.

Change-Id: I66bb0496d4998b0d6c6637c746b642b77bdbef88

1) Added an initialization of rd_tx_select_threshs[].
2) Made updating transform size counts to be consistent

Change-Id: Iaa9d6c6be825b0364c9d61a9802873d01356815c

Change-Id: I3a26baf8b2f945fea4f1aea156e60fa79f620f86

Also some unused data structures/references removed.

Change-Id: I295809e887173543e794250cb60ddaf1475ffd24

As intra coded blocks are always decoded using decode_sb_intra(), this
commmit removed the code no longer in use.

Change-Id: I09f14fa9cdc875656e8fbe245f72c8fd83b9e31e

The changing in intra coding to base on transform block, i.e. pred->
txfm->quant->dequant-itxfm->recon, made all blocks within a prediction
unit behave consistently, there is no longer a need to handle blocks
differently based on the position within a predicitn block. So this
commit simplifies the decision of transform type to be based on
prediction mode only.

Change-Id: If96cb72386f2e9186126ace88afa35ef085b6c96

This commit fixed a uninitialized value use in the intra 4x4/8x4/4x8
rate-distortion loop.

Change-Id: I5c25b3536b59e4f5fbb23cf85baf93b2ccec7d72

datatype is optional for the instruction but clang refuses it.
http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0489c/CIHIJIHC.html

It is still required when using an immediate.
http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0489c/CIHGGEEB.html

Change-Id: I0fae956c8c0fa3f97578ce80abea247f7fc88705

This commit refactors the iterative motion search for compound
inter-inter mode, to make it support all partition types including
4x4/4x8/8x4 block sizes.

Change-Id: I5f1212b0f307377291763e45c6bdc9693b5f04c8

Removal from under configure flag.
A bit  renaming

Change-Id: I2213229dfe852001dfec16b149f47c52ce88f3aa

Move 4x4/4x8/8x4 partition coding out of experimental list.

This commit fixed the unit test failure issues. It also resolved
the merge conflicts between 4x4 block level partition and iterative
motion search for comp_inter_inter.

Change-Id: I898671f0631f5ddc4f5cc68d4c62ead7de9c5a58

Added SSE2 version of variance functions for super blocks.

Change-Id: Ibeaae8771ca21c99d41dd74067574a51e97b412d

Reverts to using 128 bit LUT for the coef models rather than 48
to ease hardware implementation.

Also incorporates some cleanups including removing various
hooks to support different lookup tables based on block_type and
ref_type.

Change-Id: I54100c120cca07a2ebd3a7776bc4630fa6a153f6