GitLab

Found using the undefined behaviour sanitizer (-fsanitize=undefined).

Change-Id: I5534a97c0e38f76b38c6a53b42759b0b676e360a

check for googletest failures as well as mismatches. this greatly
reduces the error output and time to failure.

BUG=aomedia:371

Change-Id: Ic617905430a8ec39fbee2af9ce6655a8ef6796c0

Resolve the encoding failure when ec-multisymbol and rectangular
(rect-tx or var-tx) are both turned on.

Change-Id: I708ed66d907c5928adecfd2a53498566296594d6

Change-Id: I259f2983d1e9446f3e29c74f44ae583a676ce81f

promote the unsigned int calculation to uint64_t rather than int64_t for
type consistency

cherry-picked from libvpx:
47d6f16a0 get_prob(): rationalize int types

Change-Id: Ic34dee1dc707d9faf6a3ae250bfe39b60bef3438

Adds the arguments struct HandleInterModeArgs to hold arguments that
are conditional on compiled features. This means that there are no
longer #if's in the function's argument list.

Some of the array pointers that were optional arguments have been
made array members in the new struct, but not all. This is due to the
function being called with either references to arrays that are
maintained between trying different modes OR with references to
"dummy" arrays initialized to zero. The arrays that are always used
are now members of the HandleInterModeArgs struct.

Change-Id: I3076fd53c3cddf5a6d14bbe7d23a889465ed716d

+ disable squashing of commits

Change-Id: If70d57a8bc71159b633a66bb18910d3425f668ff

Change-Id: I10cb9af13d50f2e5ac471af86a4393739c28b509

Resolve compiler warnings when the two experiments are turned on.

Change-Id: I568ddd74afad6fe720e1f46d0119df08fd2f689b

Change-Id: Ie1bfece43c81ee5d149ed25c3f7fd959a8f95030

This reduces the runtime profile of pvq_search_rdo_double from 37%
to 15% and improves overall encoding speed when PVQ is enabled by ~40%.
The SIMD code is not bit accurate with the C version and introduces a
slight PSNR regression on AWCY:

  PSNR | PSNR Cb | PSNR Cr | PSNR HVS | SSIM | MS SSIM | CIEDE 2000
0.0607 |  0.1044 |     N/A |   0.0126 |  N/A | -0.0309 |        N/A

Change-Id: Ie22cebc62df2e72618305f2268668d79167860c6

Change-Id: I44842387207b19f8e0c3894d3f4e8d0646a4cafd

This is branchless on newer gcc and clang and is about 1% faster overall
at cq-level=16 frame-parallel=1.

Change-Id: I7f5608ab0f0abbc29aa3419a103addf945ea9f0a

Using 8-bit weights gives similar results as 12-bit, with only noise
level difference. Here's what 8-bit looks like compared to 12-bit:

* AWCY Objective-1-fast:
                          high latency          low latency
ALL keyframes             0.00                  0.01
Video                     0.00                  0.04

* Google sets:

All Keyframes:
lowres: 0
midres: -0.001
hdres: -0.001

Video overall:
lowres: 0
midres: -0.063
hdres: 0.026

Change-Id: Ibed6015aa7cce12fcc6f314ffde76624df4ad2a1

Offsets for the least-squares for affine motion computation
are now set at the top left corner of the current block.

Improves stability and performance a little.

Change-Id: I68ca7e74c6102502daa8ca3373af2b2dd59400c3

Disable the support of compound prediction modes for sub8x8 codking
blocks. Make the rate-distortion optimizations process account for
such constraints.

With the use 2x2 chroma prediction block, this makes the wrost case
number of inter predictors same as vp9. It affects the coding
gains by 0.35% for lowres, 0.17% for midres, and 0.08% for hdres.

The encoding speed is up by 10%.

Change-Id: Ieb2a83030676911baa403e586f1f800cbf485d81

Segmment based lossless flag is used in select transform size, this
commit fixes a bug where wrong segment_id is used in such selection.

BUG=aomedia:350

Change-Id: Ibc981c779739849bac00447155180abbd319eb28

The macro used in assert is defined under CONFIG_VAR_TX. This fixes a
build issuse when --enable-var-tx and --enable-rd-debug are both on.

Change-Id: I497fe4a8b1fa6c7b05ac2b41c97522f7bdedc0ce

Change-Id: Idf8f03052a7e21b8a273986204038545573d7962

Also supports homography models for future experiments.

Change-Id: I4510540f54133e063891ed491c95c087222f7810

The line of code is already within the scope
of an #if CONFIG_EC_MULTISYMBOL.

Change-Id: I62e28c8586f5d04a1e1be4ea5a2551d3123fde9f

USE_TXTYPE_SEARCH_FOR_SUB8X8_IN_CB4X4 macro added to turn
tx_type search on/off for sub8x8 in cb4x4 mode.

The purpose is mainly to analyze the coding gains from cb4x4
but this later can be made into a speed feature as well.

Change-Id: Ic22026c373eebba87f324689ac5686a2844315b6

Integerizes computation of the least squares for warped motion.
The model is restricted to only Affine. Affine seems easiest
to compute and integerize since it can be split into two 3-dim
least squares problems, as opposed to rotation-zoom which needs
a 4-dim least-squares problem to be solved.
The current implementation requires only one division per block.

BDRATE impact is mminimal. The upgrade to the affine model improves
coding efficiency but integerization also degrades efficiency a
little. Overall there is a net gain of about -0.07% BDRATE on
the lowres set.
BDRATE lowres: -1.113% with ----enable-warped-motion vs. without
(up from -1.044%).

Change-Id: I6b9216ac0737d76f59054293eabee48e17739ec4

- Move source list vars.
- Split source list vars into common/decoder/encoder sources.
- Move target definitions into function.
- Split targets into common/decoder/encoder targets.
- Update CMakeLists.txt to include test.cmake and call
  setup_aom_test_targets() at the appropriate time.

BUG=https://bugs.chromium.org/p/aomedia/issues/detail?id=76

Change-Id: Icd9ce67593c2de7ebd5c8ef921e31517b6d20945

The variable was later assigned value in the function.

Change-Id: I93f283a134499a050b46d9dcd6f0c0b4e8d54049

Change-Id: Ifcdd3ce2953c1ecb1d0962da412a4b5ba2cda912

--enable-lowbitdepth defines the flag CONFIG_LOWBITDEPTH, not
CONFIG_AOM_LOWBITDEPTH.

Change-Id: Ifa1c12847bee4978d08d010f4fc3601d75e59c31

Change-Id: I93bfff59810fe2d8e880012de02921ae857f300b

It only handles the realloc constraint (preserving low elements) by
serendipity, and we don't actually rely on that behavior anyway.
Meanwhile the calls may do extra copying that gets immediately clobbered
by the callers.

Cherry-pick from libvpx:
3063c3760 Remove vpx_realloc()

Change-Id: I8dfa89e4a81084b084889c27bd272fdf85184e8d

Change-Id: Id3824c09cbaae814df1d8fb029215f28e8c7a6b1

PSNR YCbCr:  -0.17%     -0.03%     -0.40%
APSNR YCbCr: -0.17%     -0.02%     -0.39%
PSNRHVS:     -0.06%
SSIM:        -0.17%
MSSSIM:      -0.07%
CIEDE2000:   -0.12%

Change-Id: I69a4b6a4e18c22c3930069396540a6fee45cb30d

Change-Id: Idd4a7ff49c594e8d1c7cb057967fe7bf75770dfb

Change-Id: Ie27b46248a3555c1182a97592d7aa7209acfae3c

Change-Id: I5a2469d10f7295f1b8a6bea64e36b8325c8dca22

Since we now require C99, this is undefined behavior.
Thanks to Luc Trudeau for the report and Alex Converse for the
suggestion on how to make the macro safe for all integer sizes.

Change-Id: I99a1342dfedb3e17a6869269be317c2ed26bfe9b

Change-Id: I222b5ce9ee1a1c1aac5620b94835967dd9a014c1

This commit enables the motion vector referencing system to use
the motion information of blocks to the bottom and right of the
collocated block. This improves the compression performance by
0.3% for lowres, midres, and hdres sets.

Change-Id: I03b3fb21f3a8698880ca9ceb945fa3e32531acdb

Add temporal motion vector referencing experiment flag.

Change-Id: I9a0fced67f4f94071bd6a38eacf9e4f311099c79

to get_binary_prob(). the only other caller mode_mv_merge_probs() does
its own test on 0.

cherry-picked from libvpx:
93c823e24 vpx_dsp/get_prob: relocate den == 0 test

Change-Id: Ie0604ad405a97ed754e4b88c6d580eb4894ea0f6

+ inline the function directly as there was only one consumer
(get_prob())

this is an attempt to reduce the amount of branches to workaround an amd
bug. this change is mildly faster or neutral across x86-64, arm.

http://support.amd.com/TechDocs/44739_12h_Rev_Gd.pdf
665 Integer Divide Instruction May Cause Unpredictable Behavior

cherry-picked from libvpx:
7481edb33 vpx_dsp/get_prob: make clip_prob branchless

Change-Id: I433059c61ce43ec5058cc16ca590d186bfa8aab5