GitLab

These large (200 KB) static stables are functionally read-only. Marking
them const moves them to the read-only section of shared libraries,
reducing memory footprint when multiple application processes are
running simultaneously.

Change-Id: I6b66ea8f512147f9762728ae01cea02a0fa71300

4:4:0 has been removed from AV1 profiles, therefore there is no need
to have these files for testing.

Change-Id: Iaea1a7bba5f3211f34dfbc5f7975841cde277cf9

This experiment of ref-adapt removes the option of COMPOUND_REFERENCE
from the frame level reference_mode setup, and hence removes 1 bit
in the frame header syntax. It now allows only SINGLE_REFERENCE or
REFERENCE_MODE_SELECT to specify in the frame header.

No noticeable coding performance impact has been observed.

Change-Id: I1bed64f927f31d396e8174a4f0c9119242bf0b5c

Make the intra mode transform block coding operation processed in
64x64 unit.

BUG=aomedia:1055

Change-Id: I7cf0f38ce7afc87febfc6b26fd663e6a3392537c

Change-Id: I48f492eefa2270c98071acc6575f75bccda7a5ce

This fixes MSVC compiler warnings

Change-Id: I0a5fecffee3c9747ec7a49748ef3dd0a7343a2bf

Change-Id: I44ad31b395c142ef2a8efa7ddcef0acf692776dd

This fixes several MSVC warnings.

Change-Id: Id94f509a56d195afb48c3f2fa71986e0d8baafb4

Change-Id: I40f4de4d58cb4958b3f3f48da4bbfc12d34dba8e

Initialize the RD_STATS variable to avoid valgrind error of
"Conditional jump or move depends on uninitialised value(s)".

BUG=aomedia:1219,aomedia:1226

Change-Id: Ie7749c8297b377750f0901d07e2c6024552a10df

Update the profile update checks correctly in aomenc to be
compliant with the new AV1 profiles.

BUG=aomedia:1220

Change-Id: I047313ec282350244986eb5c1be8c5578c2c5de3

Previously the v component assignment was being skipped.
This will make a difference with the EXT_QM experiment.

Change-Id: I079d86e98f14f933756b300092246dc800ce4c96

"ext-partition-types: Add 4:1 partitions" added a number SIMD
functions. The SAD functions introduced in that patch were not
added to the rtcd file and were not getting called.

Hash of "ext-partition-types: Add 4:1 partitions"
93c39e91

Change-Id: I47094799e27d66f74311ff0bcff23ecb7eed8a76

This experiment of frame_sign_bias is to remove the 7 bits in the
frame header that signal the sign bias value for the 7 reference
frames. Instead of explicit signaling, this experiment is to use
the frame_offset value to derive the frame sign bias. It achieves
-0.08~-0.1% coding gain on average. It has been adopted by AOM.

Change-Id: I543f2641d9a11cabfe822cd4daa8b88b938421d6

Interintra RDO is moved into motion_mode_rd instead of calling
handle_inter_mode() separately. This will save time on interp
filter and new mv search, also skips interintra modes if the base
inter predictors are not good enough.

Speedup: LL ~20% HL < 1%
Tiny performnace improvements (<0.05%).

Change-Id: If5b1ad22396df7590fbc1060e26b61734a205830

Change-Id: Ib7540dfd8a89399a4a4ffc0c221d4d710d8c8d12

Change-Id: Id4436c33dddc12d843d7975379be3c23bc73ba9c

Change-Id: I2890066b64b4fd8549a0c9a7741a7ec3274bdc3c

BUG=aomedia:1221

Change-Id: I2893ab3c6f5a2403512890ec4ea929abedb14421

The self-guided filter has now been implemented using
the intrinsics for AVX2. The corresponding speed and
correctness tests have also been added.

Note: All AVX2 functions are in synonyms_avx2.h, as
GCC produces 'ABI change' warnings if they are
included in synonyms.h.

Change-Id: I2a283a4acf8c01ee835d5edc526abc242d87ad9b

The cross_sum function has been replaced by a more optimised
version, increasing the speed of the filter by ~5%.

Change-Id: Ieb0fbe53033591919f719d0a288a55abd74ba2e4

The loop restoration parameters were previously written
at the end of each superblock, and have now been moved
to the beginning.

Change-Id: I45fc57a82d0ee340e6281f0a70d42ec0829caf9c

Follow local function naming conventions.

Change-Id: Ice840fdc03cbb0347e26326988a7520c38f77672

Replace av1_foreach_transformed_block_in_plane() function call
with partition based transform block indexing. This allows one to
make 64x64 based YUV plane coding.

BUG=aomedia:1055

Change-Id: I3a01c61697116eaa649f029808ca6b49067b540f

The decoder's memory usage has increased to where it requires at least
300MB for 1080p 4:4:4.

Change-Id: I654119b1bb569162064f8a52edce60ee63236f11

Change-Id: I2a9e16bf337a91219fc2aa10937d3cd1dfa67a71

Change-Id: I26d23cd1107b4568c6f473df373be77fddb7c752

If the intrabc motion vectors are invalid or out of range then
this patch defines deterministic decoder behaviour (disabling intrabc).
This will then allow conformance tests to be written to check this
deterministic behaviour so that we do not end up with undefined
behaviour or compatibility problems related to illegal intrabc
motion vectors.

Change-Id: I4de3e006eaf01ab07a7fbbbfd7cfe28b0b0d140c

Update the test with the hardware constraints, such as the extra delays
and wavefront support.

Change-Id: Ib6a66f2803989d20f288711dd7539ae8b192e322

Use binary flags to indicate if top-right/bottom-left referecne is
available. Pack these flags into 8-bit integers.

The table size is reduced by 8 times.

Change-Id: If026989f421f11b4fb5f54708d645f26435e5958

Consider different 1d dimensions of above and left tx blocks as
suggested in the bug report.
BDRate: -0.033%

BUG=aomedia:1168

Change-Id: I35ce915ba4164b57e4b9aad417c5ec3ea8850e91

Change-Id: Iae8bdc6916bea72683734a639efab7561e222f57

Change-Id: I8092e3bbabbcf5efb5d0c659418fdd3c6ad90f2c

when trellis quantizaiton is disabled, the b_quantizer should be used.
existing code has a lot of hard coded flag not hooked up correctly with
high level trellis quantization enable/disable flag, which causes
fp_quantizer is used. this patch is intended to remove the hard coded
flags and correctly get them from the high level control.

Change-Id: I0142aace03586d07aedea138fce08d3e72d99c32

Use cumulative range check to avoid the division in the for-loop
for base range decoding in lv-map. This makes the lv-map transform
coefficient decoding process 10% faster at the decoder end, at
high target bit-rate range.

Change-Id: I85e9646ce05f0c20c10f776ff9cdd082b16e9989

When the coding block size is below 16x16, the current tx size
search loop may end up searching 4x4 transform size multiple times.
This commit skips the redundant search steps. Tested a a few
1080p clips at 4 Mbps, the encoding speed is up by 10% for key
frame coding, and 3% for overall.

Change-Id: Ifbb27bad1ec55ef0011b1e789d995f36b3f4605e

Includes unit tests for conformance and speed.

SSSE3/CFLPredictTest:
4x4: C time = 2063 us, SIMD time = 313 us (~6.6x)
8x8: C time = 6656 us, SIMD time = 493 us (~14x)
16x16: C time = 24970 us, SIMD time = 1327 us (~19x)
32x32: C time = 59020 us, SIMD time = 5178 us (~11x)

AVX2/CFLPredictTest:
4x4: C time = 2052 us, SIMD time = 333 us (~6.2x)
8x8: C time = 6712 us, SIMD time = 513 us (~13x)
16x16: C time = 25292 us, SIMD time = 1023 us (~25x)
32x32: C time = 58994 us, SIMD time = 2828 us (~21x)

Change-Id: I08690a548be981ff10e184de468b9e0e691ee812

Change-Id: I03c6cf0b58f9e1d74c7d30d76e846be9a640dab7

Designs the sequence header syntax to support various
profiles as decided in the AV1 codec WG:

Profile 0: 4:2:0 and 4:0:0 8/10-bit
Profile 1: 4:4:4 8/10-bit
Profile 2: 4:2:0/4:0:0/4:4:4 12-bit as well as 4:2:2 8/10/12-bit

Change-Id: Iea351698280e37d65847bf75a43c5bbeba8f7cf4

This issue isn't triggered currently though, because encode_block() is
always called with enable_optimize_b = 1.

Change-Id: I1ce5a35222cb6a647418bf7e4dc3189db5aa3854