Граф коммитов

1858 Коммитов

Автор SHA1 Сообщение Дата
Yaowu Xu 12fcf74c8a Merge "Use derived variable size for memcpy" into nextgenv2 2016-10-11 16:15:43 +00:00
Yaowu Xu 4960f7c3bd Merge "Added generic SIMD support for CLPF." into nextgenv2 2016-10-11 16:05:18 +00:00
Yaowu Xu c648a9fd83 Use derived variable size for memcpy
Manually cherry-picked from aom/master:
bf2ad75a17

Change-Id: I99f05e79ec8ad35a49bc124e6dd829ccc7d9cc36
2016-10-10 17:39:29 -07:00
Steinar Midtskogen be668e92c3 Added generic SIMD support for CLPF.
Change-Id: Ie03f9a5b0a4c708a586532198d755a1e7509f149
2016-10-10 11:19:37 -07:00
David Barker 4d03d6fc6f Add sse2 forward / inverse 4x8 and 8x4 transforms
Change-Id: I89ed93fb20cf975c2b463cff58879521ceaa4163
2016-10-10 09:02:45 -07:00
Yi Luo 3a8217f21b Merge "Hybrid forward transforms 16x16 AVX2 optimization" into nextgenv2 2016-10-07 01:52:11 +00:00
Debargha Mukherjee 609453e7e4 Merge "Added sse2 inverse 8x16 and 16x8 transforms" into nextgenv2 2016-10-07 00:03:34 +00:00
Yi Luo e8e8cd8f1b Hybrid forward transforms 16x16 AVX2 optimization
- Unit tests are added for AVX2 SIMD.
- Encoder speed improvement:
  AV1 baseline and EXT_TX, three 1080p sequences at bitrate:
  800 Kbps, 2 Mbps, 6 Mbps, on i7-6700 CPU, average
  user level time reduction: 3.86%.

Change-Id: Ibbd7837ee3a831c6b1e4e471bf6c8d3fa3a19ff4
2016-10-06 15:33:15 -07:00
Peter de Rivaz 1baecfeb03 Added sse2 inverse 8x16 and 16x8 transforms
Change-Id: I43628407b11e5c8e6af4df69f2acdc67ac827834
2016-10-06 11:23:14 -07:00
Jim Bankoski ddb90bb445 decode_with_drops.sh : make sample test work for av1
Change-Id: I4175070840a6561c1cec5f5a50b64e425f3e2926
2016-10-03 11:13:13 -07:00
Yaowu Xu a7c2e5c3f5 decode_with_drop.sh: vp8->aom
Change-Id: I22dacbc2e4933a60ce7151204af9ee253990ca1f
2016-10-03 11:13:08 -07:00
Jim Bankoski 0d730f95f0 decode_to_md5_test: fixes and runs quick encode and checks decode
This test checks if there's any basic change to the bitstream or default
encoder by running an encode and checking that the md5 from the decode
doesn't change.

Any change to the default encoder or bitstream should be accompanied by
a change to the md5 in this file.

Change-Id: Ibdd5a1442296fd3e946823ec1f43e8ac4e66dd34
2016-10-03 11:12:13 -07:00
Yaowu Xu 8982e20889 decode_to_md5.sh: vp8->aom
Change-Id: I0dcb0643cf83ee99b63336df851cbca749c11b68
2016-10-03 11:11:54 -07:00
Alex Converse 57aa0f656d Merge changes Ideda50a6,Id2bced5f,If423eeb3 into nextgenv2
* changes:
  Port ANS from aom/master 25aaf40
  Refactor bitreader and bitwriter wrapper.
  Migrate aom/master ANS test from d311d02.
2016-09-29 16:43:12 +00:00
Debargha Mukherjee 4ee1f71b5c Merge "Update codec name in test enviroment to match decoder" into nextgenv2 2016-09-29 16:04:37 +00:00
Alex Converse 7fe2ae8e88 Port ANS from aom/master 25aaf40
Reconciles the following commits from aom/master to nextgenv2:
- 25aaf40bbc
- 87073de569

Change-Id: Ideda50a6ec75485cb4fa7437c69f4e58d6a2ca73
2016-09-28 12:07:00 -07:00
Alex Converse 5d33cc42c3 Migrate aom/master ANS test from d311d02.
This helps in porting entropy coder changes that happened in aom/master.

Change-Id: If423eeb3da552066cceb88227138ea61d6a20f07
(cherry picked from aom/master commit d311d02da5)
2016-09-28 12:07:00 -07:00
Peter de Rivaz 105fa6d9f2 Update codec name in test enviroment to match decoder
The codec name is defined in av1_dx_iface.c
This name needs to match kAV1Name in decode_test_driver.cc.
Otherwise the EndtoEndPSNRTest fails when built with --enable-ext-tile,
(because we need the IsAV1 function to return true.)

Change-Id: I05d5ea5b6fd4bbd49e8bcacd047fb81c27efb3b3
2016-09-28 09:11:41 -07:00
Yi Luo fda2f1b95a Merge "Add a TODO for aom_highbd_fdct16x16_1_sse2 tests" into nextgenv2 2016-09-27 22:21:54 +00:00
Alex Converse 0ad82c6edb Rename av1_ans_test to match aom/master.
Change-Id: I3b2137903a87a1f8169ff45e940575b917c26a6a
2016-09-26 13:15:41 -07:00
Alex Converse acef60bd2c Migrate bitreader to the interface from aom/master
Change-Id: I7232f9ae3d97e730f66e4b80f550192e3ef7230b
2016-09-26 12:19:11 -07:00
Alex Converse 71427df526 Merge "enums.h: Combine related #defines into packed enums." into nextgenv2 2016-09-24 00:38:53 +00:00
Urvang Joshi cb586f3ba9 enums.h: Combine related #defines into packed enums.
enums for BLOCK_SIZE, TX_SIZE and PREDICTION_MODE.

Note: These were converted to #defines earlier to save on memory:
https://chromium-review.googlesource.com/#/c/269854/

But we, instead, use attribute 'packed' (see here:
https://gcc.gnu.org/onlinedocs/gcc/Common-Type-Attributes.html#Common-Type-Attributes)
to ensure that these enums use the smallest possible integer type,
and so use smallest memory when used in structs/arrays etc.

Change-Id: If1fc136686b28847109c9f3a06f8728165e7e475
2016-09-22 09:44:51 -07:00
Alex Converse 080a2cccba Migrate bitwriter to the interface in aom/master
Change-Id: I73d46229f0feea43cbe933e51da997833cce032b
2016-09-21 11:17:08 -07:00
Yi Luo fbf5681aae Add a TODO for aom_highbd_fdct16x16_1_sse2 tests
- Here function, aom_fdct16x16_1_sse2 is mistakely tested. It can pass
  AOM_BITS_8, AOM_BITS_10, but not AOM_BITS_12. We should fix this test
  when aom_highbd_fdct16x16_1_sse2 is available.

Change-Id: I5cac6ee5404ff6d833940e1ecc34663b29d7a41c
2016-09-19 16:26:08 -07:00
Alex Converse 1ac1ae73dc Move ANS to aom_dsp.
That's where it lives in aom/master.

Change-Id: I38f405827d9c2d0b06ef5f3bfd7cadc35d5991ef
2016-09-19 09:51:27 -07:00
clang-format 67948d312d apply clang-format
Change-Id: If22018f8911d9d7ee99c2127bdfcc56e42b0e2d7
2016-09-15 16:41:21 -07:00
Jim Bankoski f7f043774b aomdec.sh : Make this test create files if needed to test decoder.
If test files don't already exist it calls aomenc to create them.

cherry-picked #ee9ac321 from aom/master

Change-Id: I0e0f33cb60b3492e9106d6c9e2c51f64f71ebb63
2016-09-09 16:39:21 -07:00
Jim Bankoski 5d105b40c3 simple_encoder: make it so we can run it in tests.
Added a limit, resolving a todo and added a limit parameter so that we
can do a very simple fast encode in 1 pass.

Change-Id: I265cd912d970d560a0b00b86e6c7ec7b6fef1e7b
2016-09-09 15:54:51 -07:00
Jim Bankoski e78a964e29 simple_decoder.sh: Support encoding in decode test scripts.
Adding AV1 input files to the test set is not feasible because the
bitstream is in constant flux. Add test input encoding and hook
it up in simple_decoder.sh to start.

cherry-picked #b591df89 from aom/master

Change-Id: Ie4c06a7c458cdc2ab003d27fb92418c77c87fc88
2016-09-09 15:49:56 -07:00
Yaowu Xu ca38a67a5c Merge "twopass_encoder: sample and test script fixed." into nextgenv2 2016-09-09 22:39:16 +00:00
Yaowu Xu 250a52ed06 Merge "set_maps: add back script and fix." into nextgenv2 2016-09-09 22:39:07 +00:00
Debargha Mukherjee 8e80f422d6 Merge "Add SSE2 versions of av1_fht8x16 and av1_fht16x8" into nextgenv2 2016-09-09 20:51:03 +00:00
Jim Bankoski a65e7beea8 twopass_encoder: sample and test script fixed.
Added a limit function and removed a todo and fixed script so that
it can actually be run on av1.

cherry-picked #1801d35d from aom/master

Change-Id: Ib8d1d1b5c7dbe0169e4e6c7d89d28801d7699c37
2016-09-09 12:38:37 -07:00
Jim Bankoski a7a3909f55 set_maps: add back script and fix.
cherry-picked #a5c5f856 from aom/master

Change-Id: Ie50a81063b5e14f4b5f3b5adcb822dba6b3ee93d
2016-09-09 12:33:09 -07:00
Jim Bankoski 19a06bccdf resize_util.sh : resize util was removed.
The app this script called was removed in this patch.
50cbe24 remove more vp8 and vp9 only code

cherry-picked #1c17dd6f from aom/master

Change-Id: Ib622eff6a3a35c5dab26908b094ace969f128c11
2016-09-09 11:51:13 -07:00
Jim Bankoski 23938a73c0 aomenc: Remove tests unsupported in av1.
Change-Id: I9379eedd577c8bfb7b82f1c996e4ee4c62ce686b
2016-09-09 18:46:57 +00:00
Yaowu Xu 3a45d57574 vp8_multi_resolution_encoder.sh: remove file
Change-Id: I3be6480b98cdde4c24b6cdfbebf362072153bcca
2016-09-09 18:46:40 +00:00
Geza Lore 1a800f6539 Add SSE2 versions of av1_fht8x16 and av1_fht16x8
Encoder speedup ~2% with ext-tx + rect-tx

Change-Id: Id56ddf102a887de31d181bde6d8ef8c4f03da945
2016-09-09 11:29:41 -07:00
Yaowu Xu 81fb4cf1ee simplify test code
Change-Id: Ib5491fb8f5dd7edf27c74abdd21b1f0a42aafd1f
2016-09-09 16:40:58 +00:00
Yaowu Xu fe24b956e9 aom_cx_set_ref: add example showing setting reference frame
Manually cherry-picked from AOM:
16944e59 aom_cx_set_ref: Example showing setting a reference frame
8f4c0ec8 examples.mk - Invalid comment fixed

Change-Id: Ifa87611561b089aebef2c132099baf265c845b10
2016-09-08 17:36:44 -07:00
James Zern 7b9407a81b s/INTERP_FILTER/InterpFilter/
this matches style guidelines and stabilizes successive runs of
clang-format across the tree. remaining types should be address in
successive commits.

Change-Id: I6ad3f69cf0a22cb9a9b895b272195f891f71170f
2016-09-09 00:32:31 +00:00
James Zern ba98061af3 av1_inv_txfm_test: fix decode-only build
fdct's are only enabled with --enable-av1-encoder

Change-Id: Iaf1dfdf713f2ecd1d215ba7ec635f353c02fa4d0
2016-09-07 16:33:35 -07:00
Yaowu Xu 01bade1064 Removed tests and data not in use
Change-Id: If688da3089ad33f18751fa2f8c46b6f5dc708bd2
2016-09-03 00:06:09 +00:00
Urvang Joshi 51dcf564b8 Merge "test_intra_pred_speed.cc : Fix visual studio build." into nextgenv2 2016-09-02 23:12:34 +00:00
Urvang Joshi 31744ec4f2 test_intra_pred_speed.cc : Fix visual studio build.
Visual studio doesn't like nested macros, apparently. This patch should
fix it.

Change-Id: Ifa56fae5be0b3dfd3fecd88a8a443e39135f96ab
2016-09-02 15:11:59 -07:00
Yaowu Xu 2ab7ff05f1 Change to use AOM copyright notice
Change-Id: I2b2b70e756b7eb9611b7b33b7d5f19b3b30e0a50
2016-09-02 19:52:03 +00:00
Yaowu Xu 890c4f2497 aomcx_set_ref -> aom_cx_set_ref
Change-Id: I60dd645451d6d65465f099a16ac855fb0b5a57a9
2016-09-02 08:54:47 -07:00
Yaowu Xu 9c01aa1b0c Change to use aom copyright notice
This minimize code differences between AOM master and nextgenv2

Change-Id: If144865bdf3ef0818e7aac11018b9e786444c550
2016-09-02 08:22:07 -07:00
Urvang Joshi 340593e530 Add ALT_INTRA experiment.
When the experiment is ON, we use Paeth predictor instead of TM
predictor.

For derf set, this gives about 0.09% improvement overall, and 0.55%
improvement if all frames are forced to be intra-only.

Also, if the EXT_INTRA experiment is also on, the improvement overall
is 0.056%, and improvement if all frames are forced to be intra-only is
0.465%.

Change-Id: Id74e107ede70a8d2107fa14fcb3f44b23a437274
2016-09-01 12:03:20 -07:00