]> granicus.if.org Git - libvpx/log
libvpx
8 years agocmake: Add partial configure.
Tom Finegan [Wed, 26 Oct 2016 18:42:42 +0000 (11:42 -0700)]
cmake: Add partial configure.

- Add minimal compiler flag testing.
- Generate aom_config.c and aom_config.h. Note: hard coded
  to generic-gnu values for now.
- Still a work in progress. This will not build anything.

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

Change-Id: Id65b42ea9f4c4f744d788660e2de7234886ce039

8 years agoaom_ports: Fix build in Xcode 8.
Tom Finegan [Mon, 7 Nov 2016 15:40:30 +0000 (07:40 -0800)]
aom_ports: Fix build in Xcode 8.

Use void casts and avoid unused/unnamed parameter warnings.

Change-Id: Id02ec2c613cb1423f693bcc56832ccd9b41d05bd

8 years agoUse block_size for max_scan_line in pvq decoding
Yaowu Xu [Mon, 7 Nov 2016 23:16:15 +0000 (15:16 -0800)]
Use block_size for max_scan_line in pvq decoding

Change-Id: I642bc205a7d2c4d472385fbeb4323e62e17984b4

8 years agoAdd transform parameter initialization
Yaowu Xu [Mon, 7 Nov 2016 22:50:35 +0000 (14:50 -0800)]
Add transform parameter initialization

The initialization of transform parameters was missing, that led to a
crash in encoder.

Change-Id: I9e35830d5f24e771c845f0d8881671d6b7228c5e

8 years agopvq: drop unused declaration
Tristan Matthews [Sat, 5 Nov 2016 01:36:51 +0000 (21:36 -0400)]
pvq: drop unused declaration

Change-Id: I5d95bb897d335dc17aa0ae5e873ba7dee46c6fda

8 years agoFix compiler warning of out-of-bound array access
Yaowu Xu [Mon, 7 Nov 2016 19:03:56 +0000 (11:03 -0800)]
Fix compiler warning of out-of-bound array access

Change-Id: I00f147cd372cedc5038708b0f23f6fae68918528

8 years agoMerge "Fix compiler warning of un-used variables" into nextgenv2
Yaowu Xu [Mon, 7 Nov 2016 22:03:12 +0000 (22:03 +0000)]
Merge "Fix compiler warning of un-used variables" into nextgenv2

8 years agoMerge "change to call fwd_txfm()" into nextgenv2
Yaowu Xu [Mon, 7 Nov 2016 22:03:00 +0000 (22:03 +0000)]
Merge "change to call fwd_txfm()" into nextgenv2

8 years agoMerge "Resolve merge issues with --enable-pvq" into nextgenv2
Yaowu Xu [Mon, 7 Nov 2016 20:31:33 +0000 (20:31 +0000)]
Merge "Resolve merge issues with --enable-pvq" into nextgenv2

8 years agoFix compiler warning of un-used variables
Yaowu Xu [Mon, 7 Nov 2016 18:45:56 +0000 (10:45 -0800)]
Fix compiler warning of un-used variables

Change-Id: I17d05bbf75a201fd010fc17e2d9bd0db8ef36d41

8 years agochange to call fwd_txfm()
Yaowu Xu [Mon, 7 Nov 2016 18:40:41 +0000 (10:40 -0800)]
change to call fwd_txfm()

The transform functions have been refactored in nextgenv2, this commit
resolves the calls in pvq patch to use this new scheme.

Change-Id: I1b56e75106a3357bb19bd7df2b4ba305eb9ed185

8 years agoResolve merge issues with --enable-pvq
Yaowu Xu [Mon, 7 Nov 2016 18:24:14 +0000 (10:24 -0800)]
Resolve merge issues with --enable-pvq

This commit resolves some compiling issues due to merge.

Change-Id: I0eef8aa36c404e185e0b0004948a49307c360d3e

8 years agoMerge "Fix bug in bicubic filter in warped_motion.c" into nextgenv2
Debargha Mukherjee [Mon, 7 Nov 2016 17:58:47 +0000 (17:58 +0000)]
Merge "Fix bug in bicubic filter in warped_motion.c" into nextgenv2

8 years agoMerge "New experiment: Perceptual Vector Quantization from Daala" into nextgenv2
Yaowu Xu [Mon, 7 Nov 2016 16:00:31 +0000 (16:00 +0000)]
Merge "New experiment: Perceptual Vector Quantization from Daala" into nextgenv2

8 years agoFix bug in bicubic filter in warped_motion.c
David Barker [Mon, 7 Nov 2016 13:47:13 +0000 (13:47 +0000)]
Fix bug in bicubic filter in warped_motion.c

Previously, do_cubic_filter would return results with the
wrong precision if the sample point was exactly aligned to
a pixel.

Change-Id: I40139f9a6701a8e72e691f37bb352f7814a7f306

8 years agoNew experiment: Perceptual Vector Quantization from Daala
Yushin Cho [Fri, 4 Nov 2016 23:36:56 +0000 (16:36 -0700)]
New experiment: Perceptual Vector Quantization from Daala

PVQ replaces the scalar quantizer and coefficient coding with a new
design originally developed in Daala. It currently depends on the
Daala entropy coder although it could be adapted to work with another
entropy coder if needed:
./configure --enable-experimental --enable-daala_ec --enable-pvq

The version of PVQ in this commit is adapted from the following
revision of Daala:
https://github.com/xiph/daala/commit/fb51c1ade6a31b668a0157d89de8f0a4493162a8

More information about PVQ:
- https://people.xiph.org/~jm/daala/pvq_demo/
- https://jmvalin.ca/papers/spie_pvq.pdf

The following files are copied as-is from Daala with minimal
adaptations, therefore we disable clang-format on those files
to make it easier to synchronize the AV1 and Daala codebases in the future:
 av1/common/generic_code.c
 av1/common/generic_code.h
 av1/common/laplace_tables.c
 av1/common/partition.c
 av1/common/partition.h
 av1/common/pvq.c
 av1/common/pvq.h
 av1/common/state.c
 av1/common/state.h
 av1/common/zigzag.h
 av1/common/zigzag16.c
 av1/common/zigzag32.c
 av1/common/zigzag4.c
 av1/common/zigzag64.c
 av1/common/zigzag8.c
 av1/decoder/decint.h
 av1/decoder/generic_decoder.c
 av1/decoder/laplace_decoder.c
 av1/decoder/pvq_decoder.c
 av1/decoder/pvq_decoder.h
 av1/encoder/daala_compat_enc.c
 av1/encoder/encint.h
 av1/encoder/generic_encoder.c
 av1/encoder/laplace_encoder.c
 av1/encoder/pvq_encoder.c
 av1/encoder/pvq_encoder.h

Known issues:
- Lossless mode is not supported, '--lossless=1' will give the same result as
'--end-usage=q --cq-level=1'.
- High bit depth is not supported by PVQ.

Change-Id: I1ae0d6517b87f4c1ccea944b2e12dc906979f25e

8 years agoMerge "Increase gm precision from 16 to 32 bit ints" into nextgenv2
Debargha Mukherjee [Sun, 6 Nov 2016 10:08:01 +0000 (10:08 +0000)]
Merge "Increase gm precision from 16 to 32 bit ints" into nextgenv2

8 years agoIncrease gm precision from 16 to 32 bit ints
Debargha Mukherjee [Thu, 3 Nov 2016 22:47:21 +0000 (15:47 -0700)]
Increase gm precision from 16 to 32 bit ints

Change-Id: I7117a6c14dc8438e4225b50bd2d3ebbaa7f850cc

8 years agoaccounting_test: fix read of uninitialized data
Tristan Matthews [Sat, 5 Nov 2016 01:09:41 +0000 (21:09 -0400)]
accounting_test: fix read of uninitialized data

Only read bits that were actually written.

Change-Id: I6d901123c319a1d92c54f511d3caa56daf882281

8 years agoboolcoder_test: fix read of uninitialized data
Tristan Matthews [Sat, 5 Nov 2016 00:59:52 +0000 (20:59 -0400)]
boolcoder_test: fix read of uninitialized data

Only read bits that were actually written.

Change-Id: Id62c52b7804cbfb401e6e7388201406bc899ea5d

8 years agoaom_ports: Silence warnings in aom_timer.h
Tom Finegan [Fri, 4 Nov 2016 16:56:25 +0000 (09:56 -0700)]
aom_ports: Silence warnings in aom_timer.h

When CONFIG_OS_SUPPORT is not enabled the aom_timer timer function
stubs cause unused parameter warnings. This comments out the arg
names and silences the warning.

Change-Id: I97bdbcbebdf081ac5cb2ffd86439028a1e672fa2

8 years agoMerge "Fix the bool coder test" into nextgenv2
Yaowu Xu [Sat, 5 Nov 2016 17:43:54 +0000 (17:43 +0000)]
Merge "Fix the bool coder test" into nextgenv2

8 years agoMerge "rdopt: clear maybe-uninitialized variable warning" into nextgenv2
James Zern [Sat, 5 Nov 2016 03:19:32 +0000 (03:19 +0000)]
Merge "rdopt: clear maybe-uninitialized variable warning" into nextgenv2

8 years agordopt: clear maybe-uninitialized variable warning
Sarah Parker [Fri, 4 Nov 2016 19:35:57 +0000 (12:35 -0700)]
rdopt: clear maybe-uninitialized variable warning

av1/encoder/rdopt.c:9533 ‘zeromv[1].as_int’ may be used
uninitialized in this function [-Wmaybe-uninitialized]
this was spurious given the logic in the if

Change-Id: I8ddfe7e46d1bf5593cc8624f05c9f181243a87d4

8 years agoFix the bool coder test
Yushin Cho [Fri, 4 Nov 2016 22:39:55 +0000 (15:39 -0700)]
Fix the bool coder test

Fix the bool coder test not to use a probability of 100%.

Change-Id: I799871cb0c48580edf0ee15a6c9931d27591ec99
(cherry picked from commit 9b79f6a3d6ea398e5d51d3d1dd69cbfb1725370e)

8 years agoMerge "Clean up write_tx_type()" into nextgenv2
Jingning Han [Fri, 4 Nov 2016 23:03:53 +0000 (23:03 +0000)]
Merge "Clean up write_tx_type()" into nextgenv2

8 years agoMerge "Refactor tx_type reader" into nextgenv2
Jingning Han [Fri, 4 Nov 2016 23:03:47 +0000 (23:03 +0000)]
Merge "Refactor tx_type reader" into nextgenv2

8 years agoMerge "Factor out common tx_type writing codes from inter/intra frame" into nextgenv2
Jingning Han [Fri, 4 Nov 2016 23:03:38 +0000 (23:03 +0000)]
Merge "Factor out common tx_type writing codes from inter/intra frame" into nextgenv2

8 years agoMerge changes Ia37f170d,Ie3082db5 into nextgenv2
Angie Chiang [Fri, 4 Nov 2016 22:58:57 +0000 (22:58 +0000)]
Merge changes Ia37f170d,Ie3082db5 into nextgenv2

* changes:
  Record YUV planes' txfm block coeff cost in handle_inter_mode()
  Separate coefficient cost of U/V planes in write_modes_b()

8 years agoMerge changes I3bc782d6,I8359e849,Iae50d0b0,Id1704d88,Ia69f13c4, ... into nextgenv2
Angie Chiang [Fri, 4 Nov 2016 22:43:30 +0000 (22:43 +0000)]
Merge changes I3bc782d6,I8359e849,Iae50d0b0,Id1704d88,Ia69f13c4, ... into nextgenv2

* changes:
  Add av1_ prefix on ###_rd_stats functions
  Use init_rd_stats() in encodeframe.c
  Add transform block coefficient cost in RD_STATS for debugging
  Add helper functions to modify RD_STATS
  Add mi_row and mi_col into mbmi to facilitate rd_debug process
  Add token cost comparison in write_modes_b()

8 years agoMerge changes I139808f4,I3d97d8db into nextgenv2
James Zern [Fri, 4 Nov 2016 22:34:11 +0000 (22:34 +0000)]
Merge changes I139808f4,I3d97d8db into nextgenv2

* changes:
  warped_motion.c: delete unused filter_4tap[]
  warped_motion.c: quiet float-conversion warnings

8 years agoRecord YUV planes' txfm block coeff cost in handle_inter_mode()
Angie Chiang [Thu, 3 Nov 2016 23:24:56 +0000 (16:24 -0700)]
Record YUV planes' txfm block coeff cost in handle_inter_mode()

Change-Id: Ia37f170d8fd961d78a751d84b9525ab7e973b81a

8 years agoAdd av1_ prefix on ###_rd_stats functions
Angie Chiang [Thu, 3 Nov 2016 22:36:18 +0000 (15:36 -0700)]
Add av1_ prefix on ###_rd_stats functions

Change-Id: I3bc782d68bcd9b52b38210eec9eecb21146fde75

8 years agoSeparate coefficient cost of U/V planes in write_modes_b()
Angie Chiang [Thu, 3 Nov 2016 23:01:03 +0000 (16:01 -0700)]
Separate coefficient cost of U/V planes in write_modes_b()

Change-Id: Ie3082db5b0fead8c322b2aeede4eff7cd723ea12

8 years agoUse init_rd_stats() in encodeframe.c
Angie Chiang [Thu, 3 Nov 2016 22:32:11 +0000 (15:32 -0700)]
Use init_rd_stats() in encodeframe.c

Change-Id: I8359e8498efd301ff81eea1d7466d0f3fec5e006

8 years agoAdd transform block coefficient cost in RD_STATS for debugging
Angie Chiang [Thu, 3 Nov 2016 19:20:58 +0000 (12:20 -0700)]
Add transform block coefficient cost in RD_STATS for debugging

Change-Id: Iae50d0b0c4f8f383ab4f91d2c1c2fa4e799c7250

8 years agoAdd helper functions to modify RD_STATS
Angie Chiang [Thu, 3 Nov 2016 18:49:15 +0000 (11:49 -0700)]
Add helper functions to modify RD_STATS

Those functions includes
init_rd_stats()
invalid_rd_stats()
merge_rd_stats()

This CL help simplify the code.

Change-Id: Id1704d883bd21a039b0478a940994ca14184ae1c

8 years agoAdd mi_row and mi_col into mbmi to facilitate rd_debug process
Angie Chiang [Thu, 3 Nov 2016 18:13:15 +0000 (11:13 -0700)]
Add mi_row and mi_col into mbmi to facilitate rd_debug process

Change-Id: Ia69f13c47f2dd34fabd220652691049166a06a68

8 years agoAdd token cost comparison in write_modes_b()
Angie Chiang [Thu, 3 Nov 2016 01:30:25 +0000 (18:30 -0700)]
Add token cost comparison in write_modes_b()

This is just partial implementation
Compare token cost of pack_mb_tokens/pack_txb_tokens with token cost
from rate-distortion loop. If there is any difference, dump out mode
info.

Change-Id: I46b373ee2522c5047f799f36baf7cec5fbc06f06

8 years agoMerge "Properly schedule the transform block recursion order" into nextgenv2
Jingning Han [Fri, 4 Nov 2016 17:53:53 +0000 (17:53 +0000)]
Merge "Properly schedule the transform block recursion order" into nextgenv2

8 years agoClean up write_tx_type()
Jingning Han [Fri, 4 Nov 2016 16:58:36 +0000 (09:58 -0700)]
Clean up write_tx_type()

Remove repeated mbmi->tx_size calls.

Change-Id: I3e4e03b69b2efffd860cc1ea34e150f4257bf081

8 years agoRefactor tx_type reader
Jingning Han [Fri, 4 Nov 2016 16:46:35 +0000 (09:46 -0700)]
Refactor tx_type reader

Factor out common codes. Remove repeated mbmi->tx_size calls.

Change-Id: Id5de35e88f1a5f16223eaa06fc2c9f69124061ef

8 years agoFactor out common tx_type writing codes from inter/intra frame
Jingning Han [Fri, 4 Nov 2016 01:31:30 +0000 (18:31 -0700)]
Factor out common tx_type writing codes from inter/intra frame

Change-Id: Id2626bd19db2504756d9a1dee709c2d08c79f771

8 years agoMerge "Remove duplicated variables in EXT_INTER" into nextgenv2
Yue Chen [Fri, 4 Nov 2016 17:11:07 +0000 (17:11 +0000)]
Merge "Remove duplicated variables in EXT_INTER" into nextgenv2

8 years agoProperly schedule the transform block recursion order
Jingning Han [Thu, 3 Nov 2016 19:47:47 +0000 (12:47 -0700)]
Properly schedule the transform block recursion order

This commit replaces the offset based block index calculation with
incremental based one. It does not change the coding statistics.

Change-Id: I3789294eb45416bd0823e773ec30f05ed41ba0dc

8 years agoFix format issue in handle_inter_mode()
Jingning Han [Fri, 4 Nov 2016 15:31:24 +0000 (08:31 -0700)]
Fix format issue in handle_inter_mode()

Change-Id: I681fd799cf46991de419cc867ccb649a6990c19d

8 years agoMerge "Further work on 64x64 fwd/inv transform support" into nextgenv2
Debargha Mukherjee [Fri, 4 Nov 2016 09:32:07 +0000 (09:32 +0000)]
Merge "Further work on 64x64 fwd/inv transform support" into nextgenv2

8 years agoMerge "Refactor: Replace rate dist sse skip by RD_STATS in VAR_TX" into nextgenv2
Angie Chiang [Fri, 4 Nov 2016 05:42:59 +0000 (05:42 +0000)]
Merge "Refactor: Replace rate dist sse skip by RD_STATS in VAR_TX" into nextgenv2

8 years agoMerge "Fix bilateral filter asan error for highbitdepth" into nextgenv2
Debargha Mukherjee [Fri, 4 Nov 2016 05:25:49 +0000 (05:25 +0000)]
Merge "Fix bilateral filter asan error for highbitdepth" into nextgenv2

8 years agowarped_motion.c: delete unused filter_4tap[]
James Zern [Fri, 4 Nov 2016 03:12:20 +0000 (20:12 -0700)]
warped_motion.c: delete unused filter_4tap[]

Change-Id: I139808f492a9e9dcac44a36237b61231ede7edc3

8 years agowarped_motion.c: quiet float-conversion warnings
James Zern [Fri, 4 Nov 2016 03:11:06 +0000 (20:11 -0700)]
warped_motion.c: quiet float-conversion warnings

Change-Id: I3d97d8db51a5a5d6b2c1cae47492b53ab37100a7

8 years agoMerge "warped_motion: Fix ubsan warning for signed integer overflow" into nextgenv2
James Zern [Fri, 4 Nov 2016 00:58:06 +0000 (00:58 +0000)]
Merge "warped_motion: Fix ubsan warning for signed integer overflow" into nextgenv2

8 years agoMerge "Fix ubsan divide by zero warning in ransac" into nextgenv2
James Zern [Fri, 4 Nov 2016 00:56:23 +0000 (00:56 +0000)]
Merge "Fix ubsan divide by zero warning in ransac" into nextgenv2

8 years agowarped_motion: Fix ubsan warning for signed integer overflow
Sarah Parker [Wed, 26 Oct 2016 19:46:03 +0000 (12:46 -0700)]
warped_motion: Fix ubsan warning for signed integer overflow

Change-Id: Ie698aa02ef56128759c71079e9bfa1af25149644

8 years agoRefactor: Replace rate dist sse skip by RD_STATS in VAR_TX
Angie Chiang [Wed, 2 Nov 2016 23:19:58 +0000 (16:19 -0700)]
Refactor: Replace rate dist sse skip by RD_STATS in VAR_TX

This is to facilitate implementation of rd_debug tool; it doesn't change
coding behavior.

Change-Id: I0eb82b31473883ba6652ed11dca09b9ec4530183

8 years agoFix bilateral filter asan error for highbitdepth
Debargha Mukherjee [Thu, 3 Nov 2016 22:14:27 +0000 (15:14 -0700)]
Fix bilateral filter asan error for highbitdepth

BUG=webm:1334

Change-Id: I5886eec0a22a8cc056e1bdb493d2faf183816656

8 years agoMerge "rdopt,global-motion: Fix -1 indexing ubsan warning" into nextgenv2
James Zern [Thu, 3 Nov 2016 22:59:33 +0000 (22:59 +0000)]
Merge "rdopt,global-motion: Fix -1 indexing ubsan warning" into nextgenv2

8 years agordopt,global-motion: Fix -1 indexing ubsan warning
Sarah Parker [Wed, 26 Oct 2016 19:48:55 +0000 (12:48 -0700)]
rdopt,global-motion: Fix -1 indexing ubsan warning

Change-Id: I1b3caf3543ab385f39f5f253c9949ad89ea5af7d

8 years agoMerge "Use TX_SIZES in intra_high_pred_fn declarations" into nextgenv2
Alex Converse [Thu, 3 Nov 2016 22:13:33 +0000 (22:13 +0000)]
Merge "Use TX_SIZES in intra_high_pred_fn declarations" into nextgenv2

8 years agoRemove duplicated variables in EXT_INTER
Yue Chen [Thu, 3 Nov 2016 22:12:42 +0000 (15:12 -0700)]
Remove duplicated variables in EXT_INTER

Introduced by merge commit 141f7a9

Change-Id: Idd68e09a6cd925d97466eabebe0e4905b5031340

8 years agoMerge "Make inline function static to fix clang compile error" into nextgenv2
Sarah Parker [Thu, 3 Nov 2016 22:09:46 +0000 (22:09 +0000)]
Merge "Make inline function static to fix clang compile error" into nextgenv2

8 years agoMerge "Don't use a TX_SIZE as a TX_TYPE" into nextgenv2
Alex Converse [Thu, 3 Nov 2016 21:44:20 +0000 (21:44 +0000)]
Merge "Don't use a TX_SIZE as a TX_TYPE" into nextgenv2

8 years agoFurther work on 64x64 fwd/inv transform support
Debargha Mukherjee [Wed, 2 Nov 2016 21:57:42 +0000 (14:57 -0700)]
Further work on 64x64 fwd/inv transform support

For higher level fwd and inv transform functions.

Change-Id: I91518250a0be7d94aada7519f6c9e7ed024574fb

8 years agoFix ubsan divide by zero warning in ransac
Sarah Parker [Wed, 26 Oct 2016 19:48:01 +0000 (12:48 -0700)]
Fix ubsan divide by zero warning in ransac

Change-Id: I8c736ff665a27ce8307fd62571b9728333756d7e

8 years agoMerge "Replace hard coded numbers with TX_SIZES macro" into nextgenv2
Debargha Mukherjee [Thu, 3 Nov 2016 19:59:09 +0000 (19:59 +0000)]
Merge "Replace hard coded numbers with TX_SIZES macro" into nextgenv2

8 years agoMerge "Fix txb_w/h use case in av1_tx_block_rd_b()" into nextgenv2
Jingning Han [Thu, 3 Nov 2016 19:31:48 +0000 (19:31 +0000)]
Merge "Fix txb_w/h use case in av1_tx_block_rd_b()" into nextgenv2

8 years agoMerge "fix build issue with --enable-delta-q" into nextgenv2
Yaowu Xu [Thu, 3 Nov 2016 18:45:57 +0000 (18:45 +0000)]
Merge "fix build issue with --enable-delta-q" into nextgenv2

8 years agoMake inline function static to fix clang compile error
Sarah Parker [Wed, 26 Oct 2016 19:50:52 +0000 (12:50 -0700)]
Make inline function static to fix clang compile error

Change-Id: I0432b8274a2764ba978dd6c4ed532fb7e4b7b519

8 years agoUse TX_SIZES in intra_high_pred_fn declarations
Alex Converse [Thu, 3 Nov 2016 17:27:40 +0000 (10:27 -0700)]
Use TX_SIZES in intra_high_pred_fn declarations

Change-Id: I078bb5244dbff153bcfab226206540ca6cebdad0

8 years agoDon't use a TX_SIZE as a TX_TYPE
Alex Converse [Thu, 3 Nov 2016 17:25:20 +0000 (10:25 -0700)]
Don't use a TX_SIZE as a TX_TYPE

Change-Id: I26b02e6578ad2d82aadfe1df2aeb84e6c11a747b

8 years agoAdd rd_debug flag
Angie Chiang [Tue, 1 Nov 2016 22:50:43 +0000 (15:50 -0700)]
Add rd_debug flag

rd_debug is a debug tool aim at finding discrepancy between rate-distortion
loop and bitstream packing.

Change-Id: I751c4121516c5e6368668229c77778880a9dcb9d

8 years agoFix txb_w/h use case in av1_tx_block_rd_b()
Jingning Han [Thu, 3 Nov 2016 16:20:08 +0000 (09:20 -0700)]
Fix txb_w/h use case in av1_tx_block_rd_b()

Match them with block_row/col index.

Change-Id: Idf0f924a093e5312b0a36b765d295e52d033eb5a

8 years agofix build issue with --enable-delta-q
Yaowu Xu [Wed, 2 Nov 2016 22:31:29 +0000 (15:31 -0700)]
fix build issue with --enable-delta-q

BUG=webm:1330

Change-Id: I120ce8ea3581018b232b19ca7ffbb07d3e99d8d0

8 years agoReplace hard coded numbers with TX_SIZES macro
Debargha Mukherjee [Thu, 3 Nov 2016 15:51:42 +0000 (08:51 -0700)]
Replace hard coded numbers with TX_SIZES macro

Replaces a couple of hard coded numbers with TX_SIZES macro
in common/reconiontra.c

Change-Id: I8a2a53ca16bc3ab51409cec340bea55292ff2dee

8 years agoMerge "Refactor recursive transform block partition search" into nextgenv2
Jingning Han [Thu, 3 Nov 2016 15:41:56 +0000 (15:41 +0000)]
Merge "Refactor recursive transform block partition search" into nextgenv2

8 years agoMerge "Make bit-stream support rectangular tx_size" into nextgenv2
Jingning Han [Thu, 3 Nov 2016 15:41:45 +0000 (15:41 +0000)]
Merge "Make bit-stream support rectangular tx_size" into nextgenv2

8 years agoMerge changes from topic 'update_dering' into nextgenv2
Yaowu Xu [Thu, 3 Nov 2016 14:03:24 +0000 (14:03 +0000)]
Merge changes from topic 'update_dering' into nextgenv2

* changes:
  Reformatting the deringing code
  Introducing OD_DERING_SIZE_LOG2 constant (3)
  Renaming deringing blockwise write-back functions to make code clearer
  Deringing refactoring: replace last_sbc with simpler dering_left flag
  Getting rid of the od_dering_in type

8 years agoMerge "Refactoring deringed block list code" into nextgenv2
Yaowu Xu [Thu, 3 Nov 2016 13:48:58 +0000 (13:48 +0000)]
Merge "Refactoring deringed block list code" into nextgenv2

8 years agoMerge "Deringing line buffer no longer depends on holding OD_DERING_VERY_LARGE" into...
Yaowu Xu [Thu, 3 Nov 2016 05:02:32 +0000 (05:02 +0000)]
Merge "Deringing line buffer no longer depends on holding OD_DERING_VERY_LARGE" into nextgenv2

8 years agoMerge "Make recursive txfm encoding process support rectangular tx_size" into nextgenv2
Jingning Han [Thu, 3 Nov 2016 04:36:55 +0000 (04:36 +0000)]
Merge "Make recursive txfm encoding process support rectangular tx_size" into nextgenv2

8 years agoMerge "Fix a merge bug between dual_filter and sub8x8mc" into nextgenv2
Jingning Han [Thu, 3 Nov 2016 01:06:38 +0000 (01:06 +0000)]
Merge "Fix a merge bug between dual_filter and sub8x8mc" into nextgenv2

8 years agoMerge "Remove redundant experimental flags from common_data.h" into nextgenv2
Jingning Han [Thu, 3 Nov 2016 01:04:45 +0000 (01:04 +0000)]
Merge "Remove redundant experimental flags from common_data.h" into nextgenv2

8 years agoRefactor recursive transform block partition search
Jingning Han [Thu, 3 Nov 2016 00:01:58 +0000 (17:01 -0700)]
Refactor recursive transform block partition search

This commit refactors the recursive transform block partition
search process to make it support rectangular transform block size
coding.

Change-Id: I0207ae40d83c7eae3cb5d460e403f470747590d3

8 years agoMake bit-stream support rectangular tx_size
Jingning Han [Wed, 2 Nov 2016 23:22:18 +0000 (16:22 -0700)]
Make bit-stream support rectangular tx_size

Allow the transform size writing, reading, and the reconstruction
process to support rectangular transform block size coding.

Change-Id: I57393c73ec60835a088d785ca838d7e3d7eb29a4

8 years agoReformatting the deringing code
Jean-Marc Valin [Wed, 2 Nov 2016 06:33:46 +0000 (02:33 -0400)]
Reformatting the deringing code

Manally removed the "clang-format off" lines. The rest is done by clang

Change-Id: I88a2028b55a541729b4e8896cdf66b544e9898bb

8 years agoRefactoring deringed block list code
Jean-Marc Valin [Tue, 18 Oct 2016 19:56:37 +0000 (15:56 -0400)]
Refactoring deringed block list code

Using a struct named dlist rather than an array named bskip. Simplified some
code.

No change in output

Change-Id: Id40d40b19b5d8f2ebafe347590fa1bb8cb80e6e1

8 years agoDeringing line buffer no longer depends on holding OD_DERING_VERY_LARGE
Jean-Marc Valin [Tue, 18 Oct 2016 07:32:52 +0000 (03:32 -0400)]
Deringing line buffer no longer depends on holding OD_DERING_VERY_LARGE

The OD_DERING_VERY_LARGE values are now explicitly copied to the buffer instead
of being read from the line buffer when we're on the edge of the frame. This
will make it possible to make the line buffer 8-bit for non-high-bitdepth.

No change in output

Change-Id: I1a4134d67ac7f8c239f08d73941405c56f01050b

8 years agoIntroducing OD_DERING_SIZE_LOG2 constant (3)
Jean-Marc Valin [Wed, 26 Oct 2016 05:26:25 +0000 (01:26 -0400)]
Introducing OD_DERING_SIZE_LOG2 constant (3)

Also cleans up the size of the deringing destination buffer.

No change in output.

Change-Id: I7fc50d862d3906ce809c1031bf0789acdf39cf34

8 years agoRenaming deringing blockwise write-back functions to make code clearer
Jean-Marc Valin [Wed, 26 Oct 2016 04:11:16 +0000 (00:11 -0400)]
Renaming deringing blockwise write-back functions to make code clearer

No change in output.

Change-Id: Ifa5df3adce9f24ef6dcd89a5f33a744bfb57194d

8 years agoDeringing refactoring: replace last_sbc with simpler dering_left flag
Jean-Marc Valin [Wed, 26 Oct 2016 03:51:26 +0000 (23:51 -0400)]
Deringing refactoring: replace last_sbc with simpler dering_left flag

No change in output.

Change-Id: I1cc2e14b2bb6c343baa7f88348c875085e5863af

8 years agoGetting rid of the od_dering_in type
Jean-Marc Valin [Tue, 25 Oct 2016 23:30:49 +0000 (19:30 -0400)]
Getting rid of the od_dering_in type

We no longer need the deringing code to be generic wrt the input depth.

No change in output.

Change-Id: I2db2beb82f1816e611cd2c0438dff217d363de33

8 years agoMerge "Remove unused get_intra/inter_scan() from scan.h" into nextgenv2
Jingning Han [Wed, 2 Nov 2016 22:50:58 +0000 (22:50 +0000)]
Merge "Remove unused get_intra/inter_scan() from scan.h" into nextgenv2

8 years agoMerge "Remove redundant config flags from get_entropy_context" into nextgenv2
Jingning Han [Wed, 2 Nov 2016 22:50:49 +0000 (22:50 +0000)]
Merge "Remove redundant config flags from get_entropy_context" into nextgenv2

8 years agoNow using a single line buffer
Jean-Marc Valin [Tue, 18 Oct 2016 00:51:53 +0000 (20:51 -0400)]
Now using a single line buffer

No change in output.

Change-Id: I4701a5517fb97889f970acfb0b44cee51c34fd95

8 years agoOnly copy data from deringed blocks to the line buffer
Jean-Marc Valin [Mon, 17 Oct 2016 21:48:12 +0000 (17:48 -0400)]
Only copy data from deringed blocks to the line buffer

No change in output

Change-Id: I6ec4a8c635337562170153585e427afd6f9d9a0f

8 years agoSplitting out 8->16 block copy code into copy_sb8_16()
Jean-Marc Valin [Mon, 17 Oct 2016 20:44:46 +0000 (16:44 -0400)]
Splitting out 8->16 block copy code into copy_sb8_16()

No change in output.

Change-Id: I4f0e37a879432e2647b3debe6a2c0c670a79dd6f

8 years agoEliminate the big superblock row buffer.
Jean-Marc Valin [Mon, 17 Oct 2016 19:38:36 +0000 (15:38 -0400)]
Eliminate the big superblock row buffer.

Now only buffering three lines across the entire frame and four lines
over the height of one superblock.

No change in output.

Change-Id: I6b99399974e197dc02f2e4ff2e60cdd7fdaa2e43

8 years agoMaking deringing buffer only one row of superblocks at a time
Jean-Marc Valin [Fri, 14 Oct 2016 17:05:14 +0000 (13:05 -0400)]
Making deringing buffer only one row of superblocks at a time

This introduces a line buffer that hold the last three lines of each original
row so that the next row can be deringed with the original input of the upper
row.

No change in output

Change-Id: I8fad3bc48745e9ce3e440289f453477a0c5442c0

8 years agoMake recursive txfm encoding process support rectangular tx_size
Jingning Han [Wed, 2 Nov 2016 22:45:07 +0000 (15:45 -0700)]
Make recursive txfm encoding process support rectangular tx_size

This commit makes the encoding process of the recursive transform
block partition support both rectangular and square transform block
sizes as the starting point. If the coding block size is rectangular,
it would allow the transform block size to start from the largest
rectangular transform size, and recursive parse to the selected
coding sizes.

Change-Id: I576628b9166565bada6a918f0a1e67849dfef4cd

8 years agoDuplicating deringing input superblock copy to make upcoming changes easier
Jean-Marc Valin [Thu, 13 Oct 2016 19:24:39 +0000 (15:24 -0400)]
Duplicating deringing input superblock copy to make upcoming changes easier

No change in output

Change-Id: Iaa06043dcc31308c83f667424e5a83c2db50ed24