By avoiding break CPU's pipeline,
this patch achieves a small encoder
speedup at the range of 0.2%~0.71%.

Change-Id: I398cb09f8eb91695e3258091ff2f82f06ab74145

explicitly disable segmentation when ref frame has different
resolution

BUG=aomedia:1205
BUG=aomedia:1223
BUG=aomedia:1256

Change-Id: I6db51116db308514d572eb465c2453403e64e1f2

This also avoids a UBsan warning of "left shift of negative values".

Change-Id: Ifb8d74218b1c0bc7b924e752442de0ba1b50a869

DC_PRED requires dividing by 'count' = block width + height.
- For square blocks this is efficient, as count is a power of 2.
- But for 1x2 and 1x4 rectangular blocks the division is inefficient.

So, we approximate this division with an integer multiply and a shift.

test_intra_pred_speed test results:

Size    Before (ms) After (ms)
4x8     221         206
8x4     195         189
8x16    121          88
16x8     84          73
16x32    63          60
32x16    68          66

Compression quality is neutral:
- Division vs 32-bit mult + shift is neutral:
https://arewecompressedyet.com/?job=ALL_rectpred_bef%402018-01-10T19%3A02%3A20.745Z&job=ALL_rectpred_aft_mult_shift%402018-01-10T19%3A04%3A21.380Z
- And 32-bit mult + shift vs 8-bit mult + shift is also neutral:
https://arewecompressedyet.com/?job=mult_shift_32bit%402018-01-16T20%3A30%3A21.418Z&job=mult_shift_08bit%402018-01-16T20%3A31%3A25.673Z

BUG=aomedia:1191

Change-Id: I2343a055bc3d85cb7be7652d7a6db1b768361be9

Change-Id: Iadaa189717787dbb82e485ff92125ee1ef591496

Add command to remove CMake outputs that will cause problems
after moving to a new directory.

BUG=aomedia:1217

Change-Id: Ia485772d907b7cb6b84d948f288dc0ce24802499

This extra entry has been introduced in 3ebb0d00
but is not needed.

Change-Id: I22801b2a0677dd5c381175f2fc6d9e86dde1594e

Change-Id: I591badb15e714e23f2c8a2c2df8514fdd75fc880

Change-Id: Ic6706ee6c39cb54ad2f746f05e80fbc6a0af88e0

The original order of operations lead to all variables upgraded to
unsigned int during computation, and enc->cnt being negative then
would cause UBsan warning.

Change-Id: I658b6f4c903472b82e178033eac2463045c42233

Can improve decoder run time by 15-20 percent.

Change-Id: I6e0ab0ce46f3620129460046cc21d99fb957adb5

It has been provisionally adopted for over a month now.

Change-Id: I717188c5b6ffa449440f253b1dede97c70fff510

Change-Id: I4f6ecb4cdd60b9ab12a403371744710be9a59f46

Use HTTP instead of HTTPS to avoid requiring CMake builds with
HTTPS support.

BUG=aomedia:1253

Change-Id: I1c130496ae3ec9ea43629117e3f3f46048a32b92

Change-Id: I4d201631e288ca0f5a0d13583707dff483eb5cac

Change-Id: If47ac212044a977b12c8df1322759d173dd29ed5

Change-Id: I50f3bec3bb67ab47a97d2731a7d2afcb4da5139a

A function to calculate chroma prediction mode info. rate cost for
intra blocks.

Change-Id: Ia71c492284675568db609dfa9ed3b5901f021027

Move the "largest tile size" value from read_tile_buffers()
into the AV1_COMMON structure, so that it can be preserved across
calls. Additionally, only reset the largest tile id once per frame,
rather than once per tile group.

BUG=aomedia:1245

Change-Id: Ia9b96047bcd4e72b1bed24f78d2914505686262c

* Remove aom_config.h include from aom_image.h
* Unconditionally include API additions in the header

Change-Id: I57e4014bb255789cc1052ed1adc38d76ca6ad9a0

This does not change the precision used for od_ec_enc_bool_q15
(used by aom_write and aom_write_bit), which means these two
paths will use different CDF precision for now.

The test that verifies that q15's code identically to CDFs
is disabled.

Change-Id: I4cd40a60d472eb58f37ac4fda1056b5dfe8b39a5

Change-Id: I69e07309b708caa59c313ae2156828d8c089cf4d

BUG=aomedia:1247

Change-Id: I53582471a5f75be26f843860914ab210f3230d5e

Creation of the testdata target was previously guarded by
ENABLE_IDE_TEST_HOSTING. This change ensures the testdata
target is always available.

BUG=aomedia:1243

Change-Id: Ief47d4c16c3718b15505e8bc63b519d0c4dbc0ae

BUG=aomedia:1243

Change-Id: Iaa3b239cf78d62432f61ad166e9c242519f6530f

Change-Id: Ifd163fd9a6eee25344f5e18e859de01181e1e48a

A function to calculate luma prediction mode info. rate cost for
intra blocks.

Change-Id: I285518dc34c0ffa58bdcd09e76463c21240336e1

encode_with_recode_loop() may encode a frame multiple times with
different q-index values. delta_q/lf_present_flag need to be
initialized every time.

Change-Id: Iee4891c9c52a06416afdb30360e377462ef85f22

Change-Id: Ib2cb09cf335352b4d0a63d3ba35a9c9f4b35b881

Change-Id: I0cd3de4a5673003f9fbe9c615e139e8133a99dcc

Change-Id: Icfdd68247543daf7884454c450d50ff95ae20712

When it is the last coeff, we don't have to search for the cost of
making it as the last signficant coeff.

Reset has_nz_tail when new eob is found.

Speed impact: ~2% improvement
PSNR: 0.01%

Change-Id: I0429d3506be30431c3e7108fb1d011930a52c419

Negative distortion is used in trellis quant process for RDcost
computation, this change avoids compiler warning on "left shift
of negative values".

Change-Id: Iaea0ffebb6c9236e779e36301fbc69b180be1f3d

Change-Id: I8bb21fbbc3366690ccc13a9b31a6726dfbfa2ff4

Initialize all the 3 planes even if the video is monochrome.
There were some unintended consequences with initializing
only the luma plane for monochrome videos, which will be
resolved subsequently.

Change-Id: Id57445e3a5a62605511220600e749e662d080b49

Skip the quantization and reconstruction process when the best
selected intra mode goes with skip mode. Tested on a few 1080p
key frame coding at mid to high bit-rate range, the encoding
speed is up by 3% - 5%.

The compression for hdres key frame coding is improved from
neutural to 0.3% gains.

Change-Id: Id6b7131e82cdb45e53db65e2b92d8ccfcf00f47e

Reduce the context model size from 3 sets to 1 set for motion
vector difference coding in opt-ref-mv.

Change-Id: Iece1e01bb44f6a445c13787aa948400b8feef118

Change-Id: I51a751cd656846e6f3db6c15e320bdc7f1c868fe

Removed compressed header related code that is not used anymore.

Change-Id: I8e4881ecf9a4e0a25bd59a456eab9815d4bdd418

The x0 parameter can now be chosen on the encoder side
independently of the dq offset.

Change-Id: I4a63bae6ae81d5a7f65c53efe1b1ecd713355b01