Use in-place filtering, like in the decoder
(see eb5794da).

Change-Id: If037ead45f5cb3461347a63e0e415954d5dcba8b

Instead of having CLPF write to an entire new frame and
copy the result back into the original frame, make the
filter able to work in-place by keeping a buffer of size
frame_width*filter_block_size and delay the write-back
by one filter_block_size row.

This reduces the cycles spent in the filter to ~75%.

Change-Id: I78ca74380c45492daa8935d08d766851edb5fbc1

Signature of aom_read_xxx function changed recently:
https://aomedia-review.googlesource.com/#/c/3520/
So, updated the palette code to call those functions with the additional
parameter.

Change-Id: I364faee2e002f2bb7ac889ee15c5f14b0d857128

The bit accounting functions aom_reader_tell() and aom_reader_tell_frac()
 return the number of bits and 1/8th bits respectively.
This patch changes the return type from ptrdiff_t which is signed to
 uint32_t which is unsigned.
The size_t type is not used since we only care about the number of bits
 or 1/8 bits per entropy coder context and we don't expect to code more
 than 512 megabits per tile.

Change-Id: I84a119d1f52829dcbdb66a92656eacca06e42b11

these aren't overly speed critical, best to leave it to the compiler. as
a side-effect this fixes Visual Studio compilation (should have been
INLINE)

Change-Id: Ic81fb5ac76bc19c61efb2f1a965c0f79e9e45ebd

This patch adds bit account infrastructure to the bit reader API.
When configured with --enable-accounting, every bit reader API
function records the number of bits necessary to decoding a symbol.
Accounting symbol entries are collected in global accounting data
structure, that can be used to understand exactly where bits are
spent (http://aomanalyzer.org). The data structure is cleared and
reused each frame to reduce memory usage. When configured without
--enable-accounting, bit accounting does not incur any runtime
overhead.

All aom_read_xxx functions now have an additional string parameter
that specifies the symbol name. By default, the ACCT_STR macro is
used (which expands to __func__). For more precise accounting,
these should be replaced with more descriptive names.

Change-Id: Ia2e1343cb842c9391b12b77272587dfbe307a56d

Purpose is to allow parsing of bitstream after a packet loss.

Change-Id: I7d1c010a465aa18914762cc1a3e61db377304c08

Change-Id: If5eb33b6b090f43ba64c82468576b89eddd872c3

read per symbol.

Change-Id: I218abaa5172b769b66dba45050381c0212602668

Cherry-picked the following commits:
e56f9639 define GET_GOT_DEFINED if not previously defined
a4b47b89 x86inc.asm: only set visibility for chromium builds
a97c83f7 Only use .text sections for aout
9679be4b Use .text instead of .rodata on macho
7e065cd5 Copy PIC handling code from x86_abi_support
71ccd103 Set 'private_extern' visibility for macho targets
1be46ef6 Expand PIC default to macho64 and respect CONFIG_PIC from libvpx
4de9641f Use libvpx defines to set name mangling rules
2d6393a9 Customize x86inc.asm for libvpx
2497356d Update x86inc.asm from x264

Change-Id: I2fe309bf7b5cea866177cec954024ced59aed789

Change-Id: Ie3f3dfe0d6ed6ab797c20887b00fc668c204f7f2

For example, loops of the form:
"for (i = 0; i < 1 + max_value; ++i) ..." or
"for (i = 0; i <= max_value; ++i) ..." are possibly infinite loops,
theoretically speaking (even if practically, they aren't).
So, compiler cannot optimize those loops.

When possible, I rewrote such loops to be finite even theoretically.

Change-Id: Ied47a24833b689c0ec011f8645cf1c01856f7c59

Change-Id: I2c504a078a7137bea6ba50c5768c1295878e9ea1

Parse and assign the tx_size with enum values in the decoding
process.

Change-Id: I85eb0f3e5d9d50c43db7694bac67befbfc8001e0

Simplify the code logic and make the codebase easier to maintain.

Change-Id: I867060614209f3d82426bb97e2e4d7d16db2097b

Make TX_4x4 as the minimum selectable transform size form bit-stream
syntax perspective. Use value 0 as the minimum supported transform
size unit for the block coding.

Change-Id: Ieb3167216a96aaddb0810a424ccaaf8ae896fec5

BDR improvements:
     PSNR  PSNRHVS SSIM  MSSSIM CIEDE2000 PSNR Cb  PSNR Cr
LL: -0.17% -0.13% -0.11% -0.12%   -0.18%   -0.19%   -0.21%
HL: -0.21% -0.14% -0.15% -0.11%   -0.37%   -0.39%   -0.52%

Change-Id: I58c00a1cc0ddfc3376644f66345e99472482a613

Change-Id: I518ad9c58973910eb0bdcb377f2d90138208c570

Change-Id: I4a6d84007bc17b89cfd8d8f2440bf2968505bd6a

It was a fairly small production optimization for VP9.

Change-Id: Ie93b474ea5b7e63384a7c0b3a56b135462d1471b

Change-Id: I544989eae45b7dda04250365c3de99f50110a76b

Change-Id: Ia09edf92bf9f7ecacc65c232ac6e656cde236634

Move computing the segmentation_probs.tree_cdf table per symbol to
 computing it only when the probabilities are updated.

Change-Id: I3826418094bbaca4ded87de5ff04d4b27c85e35a

The code that generates the raw distribution is based on a MATLAB
program by Debargha Mukherjee, and the algorithm used to quantize the
distribution comes from the ANS Toolkit by Jarek Duda.

Change-Id: Ic273f7d9e43e3ecd999e9e7e04cde57e8559375a

This brings it in line with the Daala CDFs and will make it easier to
share code.

Change-Id: Idfd2d2b33c3b9b2c4e72ce72fb3d8039013448b9

1. mem.h: improve the macro to handle n==0
2. borders_test.cc: fix a typo in

Change-Id: I24729aaf7282b2080d9a8325d481887b54341260

Change-Id: Id3cf866b147967b36f1c62e08b4a1adfb837c34c

This fixes MSVC build errors.

Change-Id: I1a05fb194284d57c03899bd597a561ece3f16a9e

Change-Id: Ib23ff963e78558a971aecc7e8b4c6184849ebf7a

Use table access to replace the enum arithmetic computations.

Change-Id: I573c301b16f30eb9fa147b3b3ec02af5cdad7444

Change-Id: Ib30ba93e86f03eb5c4705d0327d0dfc6fd9a7f72

Use table to replace the arithmetic computation for mapping between
transform block and pixel number.

Change-Id: Ica71af9891bb9ab8f0bab30794e1021066d57d64

Since diff can be negative and sum of diff can be negative too.

Change-Id: I822a4e90cc4fa06c796b7a87a776cf62a23a7621

Change-Id: I79c1b3171823376c4f6dcc72fe829e00b7bd46b4

Remove redundant variable definition.

Change-Id: Ie368c9157a3d28847b8bb5f49c6901d5d84450e8

Change-Id: Ib763971bf59859ffc457690b2f084880dc42943f

Change-Id: If27d9c12f737c20995df67ac56dc997f1eeee163

previously "vpxor" instruction was mistakenly replaced by "aomor" as
port of vpx->aom renaming change. This commit revert that change for
where vpxor instruction should be used in asm code.

Change-Id: I411d95f5d8fa5ed380c8b0d6021118b6c1e9bfb9

Use lookup table to replace the arithmetic computation for transform
block step.

Change-Id: Ie7cb33e75e3b90aa9e9b46f8d11fa1d03e1ebe85

Change-Id: I096d3f1310116ab87eaa769cdcedf400f6d62dbb